PROCEEDINGS OF€¦ · filmmakers, and feminist filmmakers can be nurtured through the festival. At the same time, the festival increases the gender self-awareness of the women who

PROCEEDINGS OF

ISERD107th INTERNATIONAL CONFERENCE

Kyoto, Japan

Organized by

Date of Event

7th-8th March 2018

In Association with

Event Co-Sponsored by

Corporate AddressIRAJ Research ForumInstitute of Research and JournalsPlot No- 161, Dharma Vihar, Khandagiri, Bhubaneswar, Odisha, IndiaMail: [email protected], www.iraj.in

PROCEEDINGS OF


Kyoto, Japan

Organized by

Date of Event

7th-8th March 2018

In Association with



PROCEEDINGS OF


Kyoto, Japan

Organized by

Date of Event

7th-8th March 2018

In Association with



Publisher: IRAJ

2018, ISERD International Conference, Kyoto, Japan

No part of this book can be reproduced in any form or by any means without prior writtenpermission of the publisher.

Type set & printed by:

R. K PrintersBhubaneswar, India

About IRAJ Research Forum (IRF):

The IRAJ Research Forum is an International non-profit academic association under

‘Peoples Empowerment Trust’ with the stated goals of promoting cooperation

among scientists, defending scientific freedom, encouraging scientific responsibility, and

supporting scientific education and science outreach for the betterment of all humanity. It is

the one of the world's largest and most prestigious general scientific society.

Objective of IRF:

To provide a world class platform to researchers to share the research findings by

organizing International/National Conferences.

To use the research output of the conference in the class room for the benefits of the

students.

To encourage researchers to identify significant research issues in identified areas, in

the field of Science, Engineering, Technology and Management.

To help dissemination of their work through publications in a journal or in the form of

conference proceedings or books.

To help them in getting feedback on their research work for improving the same and

making them more relevant and meaningful, through collective efforts.

To encourage regional and international communication and collaboration; promote

professional interaction and lifelong learning; recognize outstanding contributions of

individuals and organizations; encourage scholar researchers to pursue studies and

careers in circuit branches and its applications.

To set up, establish, maintain and manage centers of excellence for the study of /on

related subjects and discipline and also to run self supporting projects for the benefit

of needy persons, irrespective of their caste, creed or religion.

About ISERD:

International Society for Engineering Research and Development (ISERD) is a non-

profit organization that promotes the Engineering and Technology, related latest

developments and issues to be discussed and experimented through interactions amongst the

researchers and academician across the globe at a common platform in association with

The IIER.

Conference Committee

Program Chair:

Dr. P. SureshM.E, Ph.D. Professor and Controller of Examinations,Karpagam College of Engineering.,Coimbatore, India

Vitaly KlyuevUniversity of Aizu,Japan

Conference Manager:Mr. Bijan Kumar BarikMob: +91-9776047497

Conference Convener:

Miss. Manaswini Patra, ISERDMob: +91- 7539834730

Publication Head:Mr. Manas Ranjan Prusty, IRAJ, India

INTERNATIONAL ADVISORY MEMBERS

Prof. Goodarz Ahmadi,Professor, Mechanical and Aeronautical Engineering, Clarkson University, USA

Dr Chi Hieu Le,Senior Lecturer, University of Greenwich. Kent ME4 4TB. United Kingdom

PROF. (ER.) Anand NayyarDepartment of Computer Applications & I.T.KCL Institute of Management and Technology, JalandharG.T. Road, Jalandhar-144001,Punjab, India.

Prof. R. M. Khaire,Professor, Dept. Of Elex. and Telecommunication, B, V University, India

Dr.P.Suresh,Professor, Karpagam College of Engineering, Coimbatore, Tamilnadu

Mark LeesonAssociate Professor (Reader)Area of Expertise: nanoscale communications,evolutionary algorithms, network coding and communication systems

Dr. P. K. AgarwalProfessor, Deptt. of Civil Engineering, MANIT Bhopal ,Ph. D: IIT KanpurM.E: Civil Engg.IIT Roorkee, Membership: Indian Road Congress (IRC), Institute of Urban Transport (IUT)

Shahriar ShahbazpanahiIslamic Azad University,Department of Civil Engineering, Sanandaj, Kurdistan, Iran, PhD (Structural Engineering),University Putra Malaysia, Malaysia

Harun Bin SaripHead of Research and InnovationDept, UniKL-MICETDoctorate: Université de La Rochelle, FranceMember: International Society of Pharmaceutical Engineer, Singapore Chapter

Dr. Buchari LapauProfessor ,Pekanbaru Hang Tuah Institute of Health (STIKes HTP),Riau, Indonesia

Dr.Bilal Ali Yaseen Al-NassarThe World Islamic Sciences and Education University (WISE)Faculty of Business and FinanceDepartment of ManagementInformation System (MIS), Amman- Jordan

Dr. Md. Al-Amin BhuiyanAssociate ProfessorDept. of Computer EngineeringKing Faisal University, Al Ahssa 31982, Saudi Arabia

Prof. (Er.) Anand nayyarDepartment of Computer Applications & I.T.KCL Institute of Management and Technology, JalandharG.T. Road, Jalandhar-144001Punjab, India

Prof. Aleksandr Cariowinstitution or Company: West Pomeranian University ofTechnology, Szczecin

Dr. P. K. AgarwalProfessor, Deptt. of Civil Engineering, MANIT Bhopal ,Ph. D: IIT KanpurM.E: Civil Engg.IIT Roorkee, Membership: Indian Road Congress (IRC), Institute of Urban Transport (IUT)

Dr. VPS NaiduPrincipal Scientist & Assoc. Prof., MSDF Lab, FMCDCSIR - National Aerospace Laboratories, Bangalore, India

Mr. P. Sita Rama ReddyChief Scientist ,Mineral Processing Department, CSIR - Institute of Minerals & Materials TechnologyBhubaneswar, India, M.Tech. (Chem. Engg., IIT, KGP)

Dr.P.C.Srikanth,Professor & Head, E&C Dept, Malnad College of Engineering, KarnatakaSenior Member IEEE, Secretary IEEE Photonics Society,M.Tech: IIT, Kanpur, Ph.D: In IISc Photonics lab

Prof. Lalit Kumar Awasthi,Professor, Department of Computer Science & EngineeringNational Institute of Technology(NIT-Hamirpur),PhD, IIT, Roorkee, M. Tech, IIT, Delhi

Dr. Chandra Mohan V.P.Assistant Professor, Dept. of Mech. Engg., NIT Warangal,Warangal. Ph.D : Indian Institute of Technology(IIT),DelhiM.B.A: Alagappa University

Prof. I.Suneetha,Associate Professor, Dept. of ECE, AITS, Tirupati, India

Dr.s. Chandra Mohan Reddy,Assistant Professor (SG) & Head,Dept. of Electronics & Communication Engineering, JNTUA College of Engineering, Pulivendula,Ph.D,J.N.T. University Anantapur, Anantapuramu

Gurudatt Anil Kulkarni,I/C HOD E&TC Department, MARATHWADA MITRA MANDAL’S POLYTECHNIC

Pasuluri Bindu SwethaDept. Of ECE, Stanley college of Engineering & Technology for Women, Hyderabad, India

TABLE OF CONTENTS

Sl No TITLES AND AUTHORS Page No.

01. An Analysis Of Using Film As A Medium For Gender Education

Chiung-Fang Liang, Meng-Shya Rau

1-3

02. Team-Based Learning In Computer Science: A Japanese Case Study

Vitaly Klyuev

4-5

03. Type Vi Secretion System: A Syringe For Survival

Abhishek Prasad

6-12

04. In Vitro And In Vivo Study Of Bi-Layered Lidocaine And Epinephrine-ElutingPLGA/Collagen Nanofibrous Membrane

Fu-Ying Lee, Tzu-Chia Lee, Shih-Jung Liu

13-14

05. Assessment Of Website Content Based On Webometrics Ranking: Case Study OfThailand University

Tawatchai Suksida, Lalita Santiworarak

15-18

06. The Study Of Potential And Capacity Of Medical Tourism Employees On ThaiTraditional Massage And Spa Businesses In The Area Of Bang Lamphu, Bangkok

Narin Yuenthon

19-22

07. The Study On Elderly Tourist On Tourism Activities In Ranong Province

Teera Intararuang

23-26

08. Senior Tourist’s Needs Classified By Age, Towards Special Activities Provided InAccommodation In Amphawa, Samut Songkhram

Onnapat Meunpaopong

27-30

09. An Analysis Of Mobile Banking Customers For A Bank Strategy And PolicyPlanning

Behrooz Noori

31-37

10. Advantage of Make-to-Stock Strategy Based on Linear Mixed-Effect Model

Yu-Pin Liao, Shin-Kuan Chiu

38-49

11. Micro Encapsulation Of Piper Betlel. And Boesenbergia PandurataessentialOilsusing Inclusion Complexation And Their Antimicrobial Activity

Bussaba Poosri, Racha Tepsorn, Chuenjit Chancharoonpong,Suriyan Supapvanich

50-55

EDITORIAL

It is my proud privilege to welcome you all to the ISERD International Conference at

Kyoto, Japan in association with The IIER. I am happy to see the papers from all part of the

world and some of the best paper published in this proceedings. This proceeding brings out

the various Research papers from diverse areas of Science, Engineering, Technology and

Management. This platform is intended to provide a platform for researchers, educators and

professionals to present their discoveries and innovative practice and to explore future trends

and applications in the field Science and Engineering. However, this conference will also

provide a forum for dissemination of knowledge on both theoretical and applied research on

the above said area with an ultimate aim to bridge the gap between these coherent disciplines

of knowledge. Thus the forum accelerates the trend of development of technology for next

generation. Our goal is to make the Conference proceedings useful and interesting to

audiences involved in research in these areas, as well as to those involved in design,

implementation and operation, to achieve the goal.

I once again give thanks to the Institute of Research and Journals, ISERD & TheIIER

for organizing this event in Kyoto, Japan. I am sure the contributions by the authors shall add

value to the research community. I also thank all the International Advisory members and

Reviewers for making this event a Successful one.

Editor-In-ChiefDr. P. SureshM.E, Ph.D. Professor and Controller of Examinations,Karpagam College of Engineering.,Coimbatore, India.

▀ ▀ ▀

Proceedings of 107th ISERD International Conference, Kyoto, Japan, 7th -8th March 2018

1

AN ANALYSIS OF USING FILM AS A MEDIUM FOR GENDER EDUCATION

1CHIUNG-FANG LIANG, 2MENG-SHYA RAU

National Cheng Kung University, Taiwan

I. FILMS ARE RE-SHAPING THE VIEW OF THE WORLD As videos reshape our understanding of the world, films have become an important text form for our understanding of cultures or subcultures by navigating between our perspective of realities and illusions. As a text form, films not only have artistic qualities, but also serveas a rich ground for socio-cultural research and enable the universalization of cultures. In the past, in the fields of film-making training and movie productions, about 80% of movies werefilmed by men. Women had much less opportunity to use film to interpret and observe the world. II. GENDER-CENTERED WOMEN FILM FESTIVAL The only film festival in Taiwan that’s dedicated to showcase filmscreated by women, the Women Make Waves Film Festival, Taiwan (WMWFFT), has long used cinematic art to enlighten Taiwan on gender awareness. Each year, several highly experienced female filmmakers collect local and international female directors’ works to be exhibited. The WMWFFT is overseen by the Taiwan Women’s Film Association (formerlyTaipei Women Film and Video Association). The festival began in 1993 and continues to this day. It has evolved from a women’s video art exhibition in the early year to a women’s film festival and a touring festival in the recent years. It is the only film festival that tours all over the country to exhibits movies. The WMWFFT is an organization rooted in local context. It is founded, operated and supported by directors, scholars, experts, and activists who are passionate about feminist and gender/sex equality movements. Therefore, more gender equality activists, women filmmakers, and feminist filmmakers can be nurtured through the festival. At the same time, the festival increases the gender self-awareness of the women who work in the traditional filmmaking industry. The festival is the only one in the country that establishes“gender” as its theme, and for 24 years has been documenting the evolution and trends of gender issues in Taiwan. The WMWFFT reveals the content of different gender identities and awareness via the creation of films. Through various stages, we observe that

women’s imageadopts an autobiographical form to document a self-portrait using a first-person point of view. In addition, women seek to achieve gender liberation and to blur the boundaries between various gender identities through acting, costumes and disguises. Furthermore, they use female imageries, physiological experiences, and liberation of sexual desires, to tell a private narrative that belongs to women. Finally, their creativity turns to public domains such as politics, identities, labor/work, nature, heterogeneous culture, communities, immigrants, same-sex marriages, etc. III. JUSTICE IN PORNOGRAPHY - WOMEN IN SEX EDUCATION Over 24 years, films about self-empowerment, body policy, sexual liberation, queerness, and patriarchal oppressionhave never been absent at WMWFF. It’s undeniable that these stories happen all the time around the world, areinfused with issues of race, class, family, workspace power relationships etc., and present an array of wide-ranging subject matters. The heterogeneity also suggests that many voices come together as one powerful voice against sexual oppression, while it also provides us with opportunities to reflect on the insular view of sex and sexuality issues.

Fig 1. cover page of the e lectronic conference

proceeding for Women Film Festival

An Analysis of Using Film as a Medium for Gender Education


2

In 2017, the innovative theme “Justice in Pornography”put erotic films on the big screen; it audaciously overturns the male-centric perspective and boldly declares the notion of women’s sexual desires. This trend indicates that in recent years, erotic imagery is no longer a focus of men’s gaze. Women directors present women’s words and voices through a female-centric perspective and collect women’s sexual experiences and fantasies worldwide. This subversion of the male-centric perspective does not just aim forreclaiming women’s autonomous pursuit of sexual pleasuresbut goes beyond to proclaim women’s pornographic freedom. This article treats film as an extension of gender education research and selects Shu Lea Cheang, who has often been dubbed “Net Artist” and“QueerDirector” in the WMWFFT, to serve as an example to discuss how video media may deliver gender-equality-based sex education in educational settings and how to break the mold of pastgender stereotypes, and to positively and actively discuss sex, sexual desire, and physical body education. Born in Taiwan, Shu Lea Cheanggraduated from the National Taiwan University with a degree in History. After graduation, she studied and made films in New York. She is an expert in using the internet’s characteristics to weave her movies through multiple time-space settings such as multiple locations or virtual spaces on the internet. Her works thus contain infinite changeability, including changes in physical devices, images, and changes made by usersvia the internet. Through the interaction of body and multimedia, she expresses her socio-political thoughts and critique. Currently living in Europe, she’s a digital nomad and independent artist. Her works and life experiences work in combination toexamine different societies’ policies and regulations and the discriminating treatment of gender minority groups. For 20 years, her works have nearly exclusively focused on gender minority groups’ lives and rights and on the state of life of transgenders. Her works constantly challenge the antithesis of human versus technology in terms of language, culture, and systems, and explore the integration and transcendence of the human-technology relationship. Cheang challenges existing frameworks using her films. This article attempts to discuss the styles and meanings of her film series, and the groundbreaking ideas they bring to sex education within the context of gender education. In particular, her earlier series of experimental short animated films was about

sexual desires. The five films she directed (Color Scheme, Sex Fish, Sex Bowl,Coming Home,and Fingers and Kisses) have an upbeat narrative tempo and integrate experimental elements with performance artsto fully express female sexual desires and fantasies. In WMWFFT 2017, she presented two movies (IKU and FluidØ)which useda humorous style to discuss the issues around sexual desires and gender diversity with rich references to the integration of aesthetics and stream of consciousness, and broke through the limitations of space, ethnicity, nations, and ethics.

Fig 2. Shu Lea Cheang

She has worked with various art mediums and film formats, including installation,performance, net art, public art, video installation, feature length film and mobile web serial. As a net art pioneer, herBRANDON (1998-1999) was the first web art commissioned and collected by the Guggenheim museum in New York. Shehas been crafting her own film genre of new queer cinema, calling them eco-cybenoia Fresh Kill (1994), scifi cyberpunk I.K.U.(2000), scifi cyphepunk Fluidø (2017). The movie IKU discusses sexual desires from a woman’s perspective and discusses how sexual desires flows between the internet and reality. At the same time, it incorporates some contexts from Japanese erotic movies, providing an even greater room for imagination. The content of the film surrounds an elegantly-dressed, traditional Japanese beauty, Rieko, a sex cyborg who constantly changes and evolves in order to collect a variety of first hand data on sexual desires in order to produce the“IKU” microchips for sale. FluidØ is a movie set in the year 2060 when AIDS has been completely eradicated, yet the virus has evolved into a new species. ZERO GEN, the first human affected by the new virus, can produce bodily fluids that enables people to reach orgasm. The movie further demonstrates thatCheang consistently constructs themes around the possibilities regarding body, sex

An Analysis of Using Film as a Medium for Gender Education


3

autonomy, and the power and authority over one’s body. Under the forces of authorities, the depth and complexity of sex have been subverted and have

become sex for sex’s sake, where gender is no longer dichotomous, but a diverse possibility and sexual desires flows freely acrossgender boundaries.


4

TEAM-BASED LEARNING IN COMPUTER SCIENCE: A JAPANESE CASE STUDY

VITALY KLYUEV

University of Aizu, Japan

E-mail: [email protected]

Abstract - This paper discusses the specifics of team-based learning in computer science at the university in Japan. It analyzes the results of teaching an international team of graduate students. The paper pays special attention to the lessons learned to point out the educational problems. Index Terms - Coach, Computer Science, Educator, Teaching Style. I. INTRODUCTION Team based learning in higher education has gained popularity over the past years as a strategy to improve active learning by students. This type of education is actively used in medical education, engineering, computer science [1], etc. There is a large volume of publications describing training methods [2], brain-storming approaches [3], techniques to help teachers and students achieve outstanding results [4]. When teachers move to the team-based style of training students, national traditions should be taken into consideration. They may influence ways of communication between educators and students. Approaches used by teachers may need to be adjusted. In this paper, we share our experience in team-based training (coaching) Master’s students at a Japanese university. Students are measuring in computer science. Ii 2007, the University of Aizu started a new initiative: the Program for Leading Edge IT Specialists. The first results of work, research and study of the first team were introduced two years later. The first team was really international. It consisted of two Japanese students and one Russian. Nowadays, software development is not individual work. This is team work. Members of the team are people with different background. The deadlines for the development are tight. How to prepare the members to work productively in such environment? How to avoid the potential conflicts of interests? The answers to these questions are in the focus of this paper. II. PROJECT DESCRIPTION In order to understand the complexity of the project students worked on for two years, we begin with a short its description. The project title is “Summarizing tools for General Purpose Search engines”. The goal of this project is to design a prototype of the system playing role of the front-end part for the general purpose search engine. It is to download the documents returned by a search engine in response to the user

query and create small summaries automatically. Figure 1 gives an illustration of this description. Three students worked together as one team creating the prototype of the tool, testing different approaches, utilizing different methods, proposing interesting ideas. Three Master’s theses were successfully defended. It was not easy to divide the joint work into three parts. Team members paid attention to the architecture of the tool created, to the methods inside the system, and to the testing and evaluation procedures. III. TEACHING STYLE Members of the team were quite different: One Japanese student with background in computer science and poor communication skills, a Russian student with strong math background and good communication skills, and another Japanese student with background in chemistry and good English. This IT specialist initiative requires the new style of defending master theses. And the new elements are demonstration of the software developed, presenting the external reviews and the comments from the coach. To reach balanced work by team members, we introduced rotation of project manager position among the members. Every moth, they change each other in this position.

Figure 1: Web search service

Student educational background influenced their roles in the project. The student with background in

Team-Based learning in Computer Science: A Japanese case Study


5

computer science paid attention mostly to the architecture of the tool created. The student with math background was responsible for the methods inside the system. The student with background in chemistry was practicing with the testing and evaluation procedures. We pioneered in organizing so called Tea seminars when students can meet professors or scientists, talk to them and discuss the problems of the project in the relaxed atmosphere. This gives us that every student got experience in team and individual work. We run weekly group meetings. Members exposed strong internal criticism during discussions. Presentation at scientific conferences is a key part of education. The university curriculum pushed students to write reports every six months. Materials of these reports were in the basis of the scientific publications. Every student individually presented two talks at the international conferences overseas and in Japan. Team work was under control from so called external reviewers. Our reviewers are scientists from different countries: Russia, USA, and Australia. They are well known specialists in the area of information retrieval and software engineering. The reviewing was done after obtaining the first results, then in the middle of the project run, and at the time when the project was close to its finish. To get their reviews was not an easy task. Suggestions of our reviewers were very much valuable in order to improve the run of the project and to organize the work better. IV. RESULTS OBTAINED To summarize the obtained results, we should note. A working prototype of the system was successfully created. Its general Characteristics are as follows: a) It is written in Java; b) It uses the Google search engine as a back-end searcher. Special characteristics of the prototype include a) semantic-based query expansion utilizing WordNet and b) semantic-based summarization producing indicative summaries. 6 scientific papers were presented at international conferences and published in the conference proceedings. Three of these conferences are sponsored by IEEE. A list of publication is available at the Software Engineering Website of the University of Aizu [5]. Two students were offered good job at the top ranked companies in Japan. One student continued its study in Europe as PhD student. V. LESSONS LEARNED Our two year work was not easy. Let me name the problems which should be taken into account our successors: Background of each member: Disbalance in

background does not help towards the success of

the team educational process. Development of the students as independent

researchers: The participation of each student in all the key areas of problem formulation, literature survey, software specification and system evaluation is needed. We have to work more to find appropriate solution: For example, different level of English and skills in programming is a big problem

Assessment: It is not easy to evaluate individual contribution of each member of the team to ensure that each student meets expected skills. We did it through individual submission and oral presentation of project reports.

Self-organizing team: This type of team is a dream. Japanese students need pushing from outside, from their professors. In team-based education, this is the weakness and serious problem for the success of the project and educational process itself.

Job hunting: This is Japanese specifics. Beginning from the second year of studying, student were away from the university for certain periods of time. Each student was on this activity on his own schedule. We predicted this problem and did our best to push students to do as much as possible within the first year of their study

CONCLUSION This paper introduces the results of the first experience in team based learning Japanese graduate students. Students are majoring in computer science. Two years’ period of time is enough to develop a solid software application. Students are able to obtain the necessary skills to work individually and as members of one small team. At the same time, participants of this project faced the several problems connected with Japanese educational traditions. Traditional Japanese job hunting is one of them. This activity lasts for several months during the second year of study. It negatively influences the learning process. REFERENCES

[1] Robert K. Pucher, “Project Based Learning in Computer Science” The Future of Education, 4 edition, pp. 245-258, 2014.

[2] V Wallace, L Walker, “Team-based Learning” The Health Professions Educator: A Practical Guide for New and Established Faculty, Springer, pp. 67-78, 2017.

[3] Deen X. Parmelee, “Twelve tips for doing effective Team-Based Learning” Medical Teacher, vol. 32, no. 2, May 2010, DOI: 10.3109/01421590903548562.

[4] EA Gomez, D Wu, K Passerini, “Computer-supported team-based learning: The impact of motivation, enjoyment and team contributions on learning outcomes”, Computers & Education. Volume 55, Issue 1, pp. 378-390, August 2010.

[5] Software Engineering Lab, University of Aizu, on-line resource: http://web-ext.u-aizu.ac.jp/labs/is-se/ [accessed on 15.02.2018].


6

TYPE VI SECRETION SYSTEM: A SYRINGE FOR SURVIVAL

ABHISHEK PRASAD

Department of Biological Sciences, Birla Institute of Technology and Sciences, Pilani, India E-mail: [email protected]

Abstract - Bacterial Type VI Secretion Systems (T6SS) have a bacteriophage needle – like structure used to puncture target cells and deliver effectors subverting the outer membranous layer of the target cells. Within ten years since its discovery, the T6SS apparatus has been duly researched and documented. Its core components and effectors have been identified and ascribed specific functions. Utility of the T6SS in pathogenicity and inter-bacterial interactions have been studied great detail. This review covers the milestone discovery of the T6SS apparatus, an overview of its gene locus, proposed structure and core components. We also delve into the utility of T6SS apparatus in bacterial pathogenicity, sustenance, inter-bacterial interactions, communal behavior and bacterial fitness. Index Terms - Bacterial secretion system, bacterial fitness, contact dependent, Hcp, phage-like machinery, VgrG. I. INTRODUCTION A wide variety of bacteria in our surrounding are pathogenic in nature. Many co-exist in poly-microbial colonies, some maintain their niche while many others overcome competition to establish their dominance. Similarly, non-pathogenic bacteria co-exist alongside a variety of other organisms, maintaining their colonies against predators or competitors[14][25]. The prokaryotic biota has developed a variety of means to communicate between self and non-self, some of which require chemical interactions in the forms of excreted substrates and toxins[8][10].Bacterial systems have developed dedicated macromolecular secretion systems that assist in such chemical interplay. Nine secretion systems – Type I to IX (T1SS to T9SS) – have been identified till date, some of which have been well documented to have pathogenic purpose[38]. The secretion system of our interest is the Type VI Secretion System (T6SS), which has been a topic of much research due to its structural and genetic resemblance to the ‘needle – like’ bacteriophage machinery[20][29][41]. Presence of T6SS have opened new avenues inmicrobial pathogenesis and inter-bacterial competition.The present review elucidates the initial discovery, an overview of T6SS locus, structure, genes and its utility in bacterial systems. II. T6SS: INITIAL DISCOVERY The novel secretion system, Type VI Secretion System, was coined by the members of the Mekalanos laboratory in 2006[28]. Despite being named in 2006, the existence of T6SS, a novel secretion system, was hinted first in 1996 when haemolysin co-regulated protein (hcp), from Vibrio cholera, was shown to be secreted without any signal cleavage. This was unlike the other secreted proteins of V.cholera[42]. In 1997, a genetic locus was identified in Rhizobium leguminosarumbv. trifolii,which inhibited it from effectively nodulating pea[32]. This locus, later termed “imp” locus (impaired in nodulation), showed

no resemblance to any of the known genes then and its structural function was later hinted in 2003. In 1998, the gene clustericmGCDJBF in Legionella pneumophila was shown to be indispensable for killing of human macrophages. In 2003, T4bSS – related machines were identified involving IcmF-like protein associated with T4SS inL.pneumophila, and termed it IAHP (IcmF – associated homologous protein) clusters since most of the proteins in the cluster were not homologous to any known secretion system proteins. Another study in the same year showed the role of imp cluster in R.leguminosarumin temperature-dependent protein secretion. Implication of Evp protein inEdwardisellatarda as a pathogenesis factor in 2004 also hinted towards a novel secretory mechanism. Identification of IcmF homologue, SciS, in Salmonella as a regulator of intracellular bacterial levels, and the discovery of ClpV family (Clp-B like proteins) linked to bacterial – host cell interactions in 2005, made it apparent that an unknown secretion system was yet to be discovered[21][31]. In 2006, the two seminal papers published by the Mekalanos group[16][28], described extracellular export of Hcp and VgrG (3 types) from V.cholerae and P.aeruginosathat were essential for contact-dependent cytotoxicity, lacked known signal sequences and that the gene loci - VAS (Virulence – associated secretion) locus in V.cholerae and HIS-I (Hcp Secretion Island I) in P.aeruginosae - contained gene clusters distinct from those found in known protein secretion systems[21][11].The proteins encoded by imp locus and their homologues are now considered a part of T6SS.

III. T6SS LOCUS: DIVERSITY AND CONTENT Parts of the T6SS locus has been identified by numerous scientists independently since 1996, but a unified effort to document the locus started after 2006. Putative T6SS gene clusters have been identified

Type VI Secretion System: A syringe for survival


7

mainly by identifying genes connected to loss of virulence in bacteria towards target host cells, survival in infection, that are transcriptionally upregulated during an infection, or by identification of secreted proteins in the extracellular milieu [7]. Many research groups have conducted genome wide in silico studies bacterial populations to identify gene clusters to construct a bona fide T6SS locus and to check its distribution in the bacterial biota. These studies have identified T6SS gene clusters in five distinct phylogenetic groups[6][43]. However, the distribution is concentrated in Proteobacteria. Phylogenetic studies suggest that the other phyla like Acidobacteria and Planctomycetes might have gained the T6SS gene clusters through horizontal gene transfer from Proteobacteria. Over 100 gram – negative bacteria have been identified to carry T6SS gene clusters. Some species have been documented to carry more than one T6SS cluster[21], which called for a study in their inheritance and the need for multiple copies. Some studies revealed distinct evolutionary histories, as of the three T6SS gene clusters in P.aeruginosa PAO1. Further, the three gene clusters did not seem to co-regulate suggesting their independent roles in the bacterium and a possible path of horizontal gene transfer for the bacterium to have acquired them[16]. Similar studies have been carried out in Burkholderiaspecies which carry four to six clusters, Yersinia pestisand Photorhabusluminescens(four clusters). In Burkholderia spp. it was noted that although the species have four to six clusters, T6SS -5 was essential towards virulence of all the speciesin mouse model while the other clusters were dispensable. Also, T6SS -1 was important for Burkholeriathilandensis for contact dependent interactions and competitive sustenance in a mix of different bacteria[6][21][31]. Such outcomes show adaptation of different loci of T6SS for distinct purposes, like interaction with host (virulence) and other bacteria. Over the years, comparison between the identified T6SS gene clusters from various bacterial strains have encouraged researchers in outlining a set of conserved “core components” of T6SS locus.The “core components” identified listed 15 to 20 conserved open reading frames (ORFs), which were distributed among the bacterial genome. The cluster of genes varied in every bacterium in terms of gene order and orientation, but always had a conserved group of 13 essential genes[6][43]. Nine of these essential genes were identified as Hcp, VgrG, VCA0109, Vip A, Vip B, IcmF, DotU, SciN and ClpV[6][11](Fig.4). The gene products of these nine essential genes are necessary for the assembly of a working T6SS and for successful secretion of effector molecules. The T6SS clusters in the bacterial genome have either the same set of essential genes or homologues of these genes. The T6SS gene cluster contains many genes apart from the essential genes which are being studied to understand their role as structural or secretory components of the

T6SS [9][31] While genome studies have identified T6SS gene loci in almost all pathogenic bacterial strains, similar studies have identified the presence of T6SS in prokaryoteshaving neither pathogenic nor symbiotic relations with any host – for example in Myxococcusxanthusand Pseudomonas putida, which hints towards a more versatile utility of T6SS in bacteria. IV. T6SSPHAGE – LIKE MACHINERY The T6SS structure was one of the major reasons for the interest of the research fraternity in this secretion system. The T6SS has a structural similarity to that of the bacteriophage “needle” that allows the T6SS containing bacteria to subvert the cell wall of the host to deliver substrates or toxins directly into the host cell, very similar to a bacteriophage infecting a bacterial host with their genomic material [36][]. The initial idea of the T6SS structure was documented through X-ray crystallography, electron microscopyand single particle analysis of Hcp1 secreted by P.aeruginosa in 2006[28].The Hcp1 proteins form ahexameric ring structures with an outer diameter of 8.5nm and an internal diameter of 4 nm (Fig 2). The Hcp1 and were shown to stack onto each other, which led to a proposition that Hcp1 proteins may assemble to form cylindrical pipe –like conduitsfor substrate transport out of the bacterial cell. Further studies revealed that Hcp1 rings could self-assemble into nanotubes in vitro, with ring-ring interfaces stabilized by di-sulfide bonds. The Hcp1 structure was very similar to the λ – phage tail protein, while the internal and outer dimensions of the Hcp1 nanotube were nearly identical to the tail tube of bacteriophage T4[11][15][41][44]. Genetic comparison of the Hcp1 gene with that of the gene product 19 (gp19) of T4 showed sequence similarities of Hcp to gp19which polymerizes to create the phage tail tube in T4. Similar homology of other T6SS essential genes to bacteriophage parts built the notion of a phage-like “needle” structure for T6SS. Like Hcp1, VgrG (Valine – Glycine repeat protein G)proteins have been identified as secretory proteins of T6SS due to their presence in the culture supernatants of several bacteria. VgrG proteins have been shown to assemble as trimers and have structural homology to the hub proteins (gp27, gp5) of the bacteriophage T4 base-plate. The amino-terminal of VgrG proteins are similar to gp27 while the central domain of resembles that of gp5. The gp27 and gp5 proteins in T4 associate in a stoichiometric ratio of 3:3 to form a torch-like structure, forming the tail spike complex of the bacteriophage base-plate. The gp27 trimer assembles into a cylinder with internal diameter of 3.0 nm and the gp5 C-terminal domains arrange to form a twisted β-helical prism at the end of the cylinder. The prism narrows down from 3.3 nm at the base to 2.5 nm at the tip, resulting in a needle – like structure at the base plate (Fig 3). V.choleraeVgrG proteins form trimers and may have a similar



8

needle-like arrangement and function as a mode to puncture into the cell wall of the host organism and act as a syringe to inject T6SS substrates/toxins into the host[19][20][37].V.choleraT6SS proteins VipA and VipB are similar the T4 tail – sheath proteins, which encase the bacteriophage tail tube and on contraction pushes the tail tube through the outer membrane of host bacteria. VipA/VipB form a cog-wheel like tubular complex with an internal diameter of 10.0 nm which can, in theory, accommodate the 8.5 nm thick Hcp nanotube[7][41]. Such an assumption allows the VipA/VipB protein complex to act as a contractile - tail sheath which can contract to push the encased Hcp nanotube through the bacterial outer membrane exposing Hcp and VgrG to exta-cellular milieu[24]. It may also help in thrusting the needle-like Hcp – VgrG structure through the outer membrane of the host cell[12][41]. VCA0109 gene shares sequence homology to gp25 of bacteriophage T4, which forms the part of phage tail base plate. Similar to other bacterial secretion systems IcmF and DotU proteins have shown to interact with each other. IcmF homologues have been documented to be give structural support to T4SS in Legionella pneumophila. DotU homologues are also present in this bacteria which interact with the T4SS and IcmF, providing structural support to T4SS. This, and the study of IcmF and DotU interaction in T6SS systems, and their presence in the inner membrane may hint towards a similar structural role of IcmF and DotU for T6SS too. A similar interaction of IcmF and SciN, a lipoprotein, in a yeast two-hybrid assay has hinted at a structural role of SciN in T6SS, alongside IcmF and DotU[2][5][43]. A membrane core complex[12], which acts as a T6SS docking site involves these three protein complexes.ClpV protein, discovered in 2005, belongs to the ClpB family, which use ATP to unfold protein substrates. It may have a conserved ATPase role in T6SS due to its similarity to the ATPase components of other bacterial secretion systems. However, S.entericaClpV proteins failed to unfold aggregated proteins.Thus, while ClpV, which forms a hexameric ring, may deliver unfolded protein effectors to Hcp nanotubes, it may alternately be responsible for ATP–driven remodeling of VipA/VipB and further assembly of the T6SS machinery instead [5][16]. As such, T6SS is thought to have an inverted phage tail – like setup on the surface of the bacterium. This model is consistent with the structural studies of T6SS, its contact-dependent effects, and that VgrG gains access to the cytoplasm of the targeted cells[33][41]. A study of Tse proteins, the most documented T6SS effectors, presented the extent of penetration of the phage-like T6SS into a gram-negative host cell to reach the periplasm only, where the Tse proteins are injected. Further transport of the Tse proteins were subjected to an independent pathway [30][33]. Recent research has also increased our understanding of the T6SS assembly involving base-plate positioning,

anchorage of T6SS in the bacterial membrane, elongation of hcp nanotube and contraction of VipA/VipB sheath and its disassembly[12].

Fig. 1. Schematic Representation of T4 Bacteriophage (A) and

Predicted T6SS structure (B) [31].

Fig.2. P.aeruginosaHcp1 structure. (Inset scale bar – 10nm).

Crystal structure of Hcp1 forms rings with an internal diameter of 4nm and and outer diameter of 8.5 nm. Model reproduced

from[28]

Fig. 3. (a) Domain organization of VgrG. The C-terminal

domain may act as an effector of T6SS while the N- terminal domain has structural homology to gp27 of T4 bacteriophage and central domain has structural homology to gp5 –C of T4. VgrG proteins assemble into trimeric structure and may have resemblance to the assembly of gp27 (b) and gp5-C (c) to form

the hub at the base-plate of T6SS structure[7]. V. T6SS EFFECTORS: PART OR INDEPENDENT By definition, effector proteins have dedicated secretion systems and are similar to eukaryotic proteins with regulatory, signaling and enzymatic activities[7]. Many studies have been conducted to understand the origins of the effector proteins for T6SS – whether the essential proteins may act as effectors or independent proteins are expressed as effectors for T6SS[39]. The T6SS locus has many genes which do not belong to the core components. A few of these genes encode proteins identified as effectors of T6SS[12]. One such gene, EvpP in E.tarda, has been shown to interact with Hcp and retrieved



9

from infected fish species[43]. Tse (Type VI secretion exported) proteinsfrom P.aeruginosa have been characterized as T6SS effectors too. Tse2 which is targeted as a toxin component of a toxin-immunity system, has been shown to cause arrested growth in bacterial host cells lacking Tsi2 (Type VI secretion immunity 2) protein, the cognate immunity protein to Tse2. The other two members of Tse protein group – Tse 1 and Tse 3 –are bacteriolytic enzymes that degrade the peptidoglycan layer in bacterial host through amidase and muramidase activity [33]. These three proteins are coded in HIS – I (Hcp secretion island – I) of P.aeruginosa, and are non-essential proteins. Such results hint towards possible effector encoding functions of non-essential genes in T6SS. Bacteriocins are molecules released by bacteria as narrow range antimicrobials. Intriguingly, several T6SSs have genes encoding proteins similar to bacteriocins. Sequence analysis identified aVgrG protein having bacteriocin-like C-terminal domain present in SPI -21 T6SS gene cluster of S.enterica. Similarly, a gene encoding a pyocin-like fused Hcp has been identified in uropathogenicEscherichia coli [34]. Other than bacteriocin-like modifications, VgrG proteins have C-terminal domains that resemble known virulence factors or eukaryotic domains. The C-terminal domain of V.cholerae VgrG1 protein has homology to actin cross-linking domain (ACD) of the RtxA toxin. Experimental analysis has shown VgrG1 dependent actin cross-linking of macrophages [10]. Sequence study of AeromonashydrophilaVgrG C-terminal has shown homology to eukaryotic lysosomal cathepsin D protein, suggesting probable host – cell apoptosis under exposure to VgrG. An ectopic production of A.hydrophila Hcp in HeLa cells amplified the rate of caspase-3 mediatedby apoptosis[7]. These studies have hinted at “dual” activity of essential T6SS proteins like Hcp and VgrG, having both structural and effector functions in T6SS. Recent studies have identified numerous effectors based on their targets and interactions with core components of T6SS [12][36]. VI. T6SS REGUATION: A COMPLICATION A variety of regulatory pathways have been identified in different T6SS carrying bacterial population, both transcriptional [22][37] and post-transcriptional. A common positive transcriptional regulator being the AraC family or sigma 54 has always been identified with regard to T6SS. Upregulation of most T6SS clusters is dependent on contact between bacteria and host, or intracellular growth of bacteria in host during infection process. A post- transcriptional regulatory mechanism identified in P.aeruginosainvolves a threonine – phosphorylation signaling cascade controlled by two proteins – PpkA (a serine/threonine(S/T) kinase) and PppA (a S/T

phosphatase) – encoded within the HIS –I cluster [17].The T6SS genes are mostly induced in vivo during infection. Several two-component systems have been identified that might be involved in induction/repression of T6SS, but their specificity to species suggests further research in this area. VII. T6SS UTILITY: A VERSATILE TOOL The T6SS has been implicated as a virulence factor [28][29][35], an inter-bacterial mediator, a defense measure, and as a mode to facilitate community behavior in microbiota[1][4][12][14][18][23].In many pathogens T6SS mutation correlated to virulence defects. T6SS has been shown to facilitate adherence,host-cell invasion,intracellular growth, cytotoxicity etc. E.tardaevp mutants show lower replication rates than wild types, such that infections are notfatal. In A. hydrophila, T6SS clusters were essential to inhibit phagocytic activity of macrophages. F.tularensisVgrG is essential for phagosome biogenesis regulation and for intracellular replications. Cytotoxicity of V.cholerae towards macrophages has been associated with actin cross-linking activity of VgrG1. In P.aeruginosa, mutation in HIS-I cluster decreases/ abolishes virulence in rat models [7]. While, T6SS has been identified to control contact-dependent virulence in some species, recent studies have pointedto have a role in sustenance of bacteria by decreasing its virulence. Studies with Salmonella entericaserovar Typhimurium have identified the role of SciS gene (an ortholog of icmF) in limiting replication of macrophages and a decrease of the pathogen’s virulence in a mouse model. Such an activity contributes to long–term sustenance of bacteria in host [31]. A similar observation was made when icmF or hcp mutation in Helicobacter hepaticus showed an increase in its virulence in the gastrointestinal tracts of rodents. Furthermore, H.hepaticus T6SS mutants induced higher inflammatory responses than wild –types, leading to a conclusion that T6SS may decrease inflammatory responses in host to provide a sustainable relationship between bacterium and host [8]. T6SS has also been characterized as a means for bacterial competition among the prokaryotic biota.Burkhoderia thialndensis T6SS -1 is involved in inter-species interactions with other proteobacteria. T6SS-1 mutants showed arrested growth when sensitized with P.putida, P.fluorescensetc, while the wild–type of B.thilandensisgrewin the same environment [35].P.aeruginosa Tse2 caused arrested growth in bacteria not expressing Tsi2 [30]. In predator-prey experiments,treatment with V.cholerae sp. strain V52 decreased the population of culturableE.coliby a 100,000- fold. Similar virulence of V.choleraetowards P.aeruginosawas not seen, indicating limited range of antimicrobial effects of T6SS.This observation was T6SS contact



10

dependent[31].Lysobacterenzymogens, a pathogen of fungal hosts, has a reduced ability to colonize fungal cells under hcp mutation[27]. Prevalence of T6SS in bacteria which are likely to encounter a variety of competing microorganisms, and presence of bacteriocin homologues in their T6SS clusters indicate specialization of T6SS in these bacteria to improve competitive fitness. Such a specialization involves specificity in the system that comes from the variety of effector molecules and the necessity of contact for action of T6SS. T6SS has also been documented to providehost-specificity in bacteria too. Nitrogen-fixing R.leguminosarum strain RBL5523 forms root nodules in both vetch and pea plants, but shows nitrogen-fixation activity in vetch only. Deletion of impJ allowed RBL5523 to form nitrogen-fixation competent root nodules in pea too showing involvement of T6SS in identifying the host organism[32].T6SS as means of defense mechanism have been shown in P.aeruginosa. P.aeruginosa can perceive T6SS attacks from neighbouring cells, and retaliate with a T6SS mediated attack, or “T6SS dueling”. In defense based (targeted firing) application of T6SS, as in P.aeruginosa, the bacterium gets attacked before it attacks back. Such bacteria do not target T6SS – defective bacteria, in contrast to offensive bacteria likeV.cholerae[4]. Community and social behavior is also shown to be regulated by T6SS. In Proteus mirabilisidsB (identification of self B) protein, a VgrG homolog, regulates recognition of self- and non-self, causing the formation of distinct boundaries between various strains ofP.mirabiliswhen grown on an agar medium [1].Similar phenomenon of communal growth was documented inMyxococcusxanthus. Myxobacteriaform fruiting bodies under stressful

environmental conditions. On deletion of an M. xanthusT6SS locus it was observed that the bacteria was unable to produce mature fruiting bodies although its growth or motility was unaffected[18]. Disruption in biofilm formation or inability to form biofilms have been noted in P.aeruginosa, Vibrio parahaemolyticus, E.coli and P.fluorescens, carrying mutant T6SS core genes[9][13]. Recent research showing involvement of T6SS in maintaining gut microflora, and quorum sensing hint at a holistic role in improving bacterial fitness in the environment. Study of commensal bacteria, like Bacteroidesfragilis,documented the role of T6SS in defense against predators and also in controlling microbial composition in host gut, thus influencing host health. B. fragilisutilizes T6SS to antagonize numerous bacteroidale species isolated from the human gut. The non-toxicogenic B.fragilis also limits acquisition of pathogenic enterotoxicogenicB.fragilis, thus controlling the microflora of the gut[14][40]. Such results hints towards probable research in engineering probiotic commensal species that use T6SS to specifically target certain enteric pathogens[3]. VIII. T6SS GENE NOMENCLATURE Identification and characterization of a substantial number of T6SS gene clusters led to a problem of gene nomenclature following which a unified nomenclature was proposed in 2007 [7]. We have used the older version of nomenclature, based on protein discoveries, to facilitate the reader to comprehend research conducted on T6SS before 2007. Fig.4 lists down the new nomenclature.

Fig.4. List of T6SS genes annotated a unified nomenclature. All homologues discovered in various species have been grouped under an individual Tss gene. The table is interpreted as follows: TssA – SciA,VasJ/L, EvpK, ImpA; TssD – Hcp, SciK/M, EvpC; TssI –

VgrG, VgrS.[7] – Supplementary



11

CONCLUSION There has been vast research done in characterizingthe T6SS structure, secretory elements, functions and regulation since 2006. Its ability to mediate bacterial interactions, virulence and bacterial fitness, and detailed pathways and motifs controlling T6SS activation/inhibition are yet to be completely comprehended. Their involvement in bacterial fitness in a microbiota also hints at the involvement of T6SS in poly microbial diseases. While active study on T6SS mediated bacterial control is underway, our team of researchers at the Dept. of Biological Science, Birla Institute of Technology and Sciences (BITS)Pilani, Rajasthan, India, are exploringthe impacts of T6SS on pathogenicity and plant growth regulation characteristics ofEnterobacter cloacae, a member of plant growth regulating rhizobacteria (PGPR) responsible for improvingplant fitness against salt stress in Triticumaestivum. ACKNOWLEDGEMENTS I thank Dr. PrabhatNathJha (Head of Department, Biological Sciences, BITSPilani, India) and Shahid Khan (BITS Pilani University) for their insights and critical review of this manuscript. REFERENCES [1] Alteri C. J., Himpsl S. D., Pickens S. R., Lindner J. R., Zora J.

S., Miller J. E., et al., "Multicellular bacteria deploy the type VI secretion system to preemptively strike neighboring cells.," PLoS Pathog, vol. 9, 2013..

[2] Aschtgen MS, Bernard CS, De Bentzmann S, Lloubes R and Cascales E, "SciN is an outer membrane lipoprotein required for type VI secretion in enteroaggregative Escherichia coli," J Bacteriol, vol. 190, pp. 7523-7531, 2008.

[3] Baron C and Coombes B, "Targeting bacterial secretion systems: benefits of disarmament in the microcosm," Infect Disord Drug Targets, vol. 7, pp. 19-27, 2007.

[4] Basler M., Ho B. T. and and Mekalanos J. J., "Tit-for-tat: type VI secretion system counterattack during bacterial cell-cell interactions," Cell, vol. 152, pp. 884-894, 2013

[5] Bönemann G., Pietrosiuk A. and and Mogk A., "Tubules and donuts: A type VI secretion story," Mol. Microbiol, vol. 76, pp. 815-821, 2010.

[6] Boyer F., Fichant G., Berthod J., Vandenbrouck Y. I. and Attree, "Dissecting the bacterial type VI secretion system by a genome wide in silico analysis: What can be learned from available microbial genomic resources?," BMC Genomics, vol. 10, 2009.

[7] Cascales Eric, "The type VI secretion toolkit," EMBO Reports, 2008.

[8] Chow J. and Mazmanian S. K.,"A pathobiont of the microbiota balances host colonization and intestinal inflammation," Cell Host Microbe, vol. 7, 2010.

[9] Christopher J. Alteri and Harry L. T.Mobely, “The Versatile Type VI Secretion System,” Microbiol Spectr, 2016.

[10] Efstathios Giaouris, Even Heir, Michaël Desvaux, Michel Hébraud, TrondMøretrø, Solveig Langsrud, et al., “ Intra- and inter-species interactions within biofilms of important foodborne bacteria pathogen,” Frontiers in Microbiology, 2015.

[11] Filloux Alain, Hachani Abderrahman and Bleves Sophie, "The bacterial type VI secretion machine: yet another player for protein transport across membranes," Microbiology, vol. 154, pp. 1570-1583, 2008.

[12] Gallique M, Bouteiller M and and Merieau A, "The Type VI Secretion System: A Dynamic System for Bacterial Communication?," Front. Microbiol, vol. 8, 2017.

[13] Gallique M., Decoin V., Barbey C., Rosay T., Feuilloley M. G. J., Orange N., et al., "Contribution of the Pseudomonas fluorescens MFE01 type VI secretion system to biofilm formation.," PLoS ONE, vol. 12, 2017.

[14] Hecht AL, Casterline BW, ZM Earley, Goo YA, Goodlett DR and Bubeck Wardenburg J., "Strain competition restricts colonization of an enteric pathogen and prevents colitis," EMBO Rep., vol. 17, no. 9, pp. 1281-1291, 2016.

[15] Jobichen C., Chakraborty S., Li M., Zheng J., Joseph L., Mok Y.K., Leung K. Y. and and Sivaraman J., "Structural basis for the secretion of EvpC: A key type VI secretion system protein from Edwardsiella tarda.," PLoS ONE , Pubished online, 2010.

[16] Joseph D. Mougous, Cuff Marianne E., Stefan Raunser, Aimee Shen, Min Zhou, Casey A. Gifford and et al., "A Virulence Locus of Pseudomonas aeruginosa Encodes a Protein Secretion Apparatus," Science, vol. 312, no. doi: 10.1126/science.1128393, pp. 1526-1530, 2006.

[17] Joseph D. Moughous, Gifford CA,Ramsdell TL, Mekalanos JJ,”Threonine phosphorylation post-trasncriptionally regulates protein secretion in Pseudommonas aeruginosa,” Nat Cell Bio, 2007.

[18] Konovalova A., Petters T. and and Sogaard-Andersen L., "Extracellular biology of Myxococcus xanthus," Fed. Eur. Microbiol. Soc, vol. 34, pp. 89-106, 2010.

[19] Leiman P. G., Basler M., Ramagopal U. A., Bonanno J. B., Sauder J.M. and Pukatzki et al., "Type VI secretion apparatus and phage tail-associated protein complexes share a common evolutionary origin.," Proc. Natl. Acad. Sci. USA, vol. 106, pp. 4154-4159, 2009.

[20] Leiman P. G., Chipman P. R., Kostyuchenko V. A., Mesyanzhinov V. V. and and Rossmann M. G., "Three-dimensional rearrangement of proteins in the tail of bacteriophage T4 on infection of its host.," Cell, vol. 118, pp. 419-429, 2004.

[21] Lewis EH Bingle, Christopher M B and Mark J Pallen, "Type VI secretion: a beginner’s guide," ELSEVIER, vol. 11, 2008.

[22] Martina Lazzaro, Mario F. Feldman and Eleonora García Véscovia, "A Transcriptional Regulatory Mechanism Finely Tunes the Firing of Type VI Secretion System in Response to Bacterial Enemies," vol. 8, no. 4, 2017.

[23] Martina S Corsi and Manuela R, "Close encounters of the type-six kind: injected bacterial toxins modulate gut microbial composition," EMBL report, 2016.

[24] Maximilian Brackmann, Jing Wang & Marek Basler, "Type VI secretion system sheath inter-subunit interactions modulate its contraction," EMBO Reports, 2017.

[25] Michael E. Hibbing, Clay F., Matthew R. P. and S. Brook Peterson, “Bacterial competition: survivingand thriving in the microbial jungle,” Nature Reviews, vol. 8, doi: 10.1038/nrmicro2259, pp. 15-25, 2010.

[26] Mobley Christopher J. Alteri and Harry L.T., "The Versatile Type VI Secretion System," Microbiol Spectr, vol. 4, no. doi:10.1128/microbiolspec., 2016.

[27] Patel N., Blackmoore M., Hillman B. and and Kobayashi D., "Evidence for the role of Type VI secretion during Lysobacter enzymogenes pathogenesis of fungal hosts.," Phytopathology, vol. 99, pp. 100-101, 2009.

[28] Pukatzki S., Ma A. T., Sturtevant D., Krastins B., Sarracino D., Nelson W. C. and Heidelberg et al., "Identification of a conserved bacterial protein secretion system in Vibrio cholerae using the Dictyostelium host model system," Proc. Natl. Acad. Sci. USA, vol. 103, pp. 1528-1533, 2006.

[29] Pukatzki S., T. Ma A., R. A. T. Sturtevant D. and Mekalanos J. J., "Type VI secretion system translocates a phage tail spike-like protein into target cells where it cross-links actin.," Proc. Natl. Acad. Sci. USA, vol. 104, pp. 15508-15513, 2007.

[30] Rachel D. Hood, Pragya S., Fo Sheng H., Tüzün G., Mike A. Carl, Rex R. S. Trinidad et al., "A Type VI Secretion System of Pseudomonas aeruginosa Targets a Toxin to Bacteria," Cell Host Microbe, vol. 7 pp. 25-37, 2010.



12

[31] Records Angela R., "The Type VI Secretion System: A Multipurpose Delivery System with a Phage-Like Machinery," The American Phytopathological Society, vol. 247, no. 7, pp. 751 - 757, 2011.

[32] Roest H. P., Mulders I. H. M., Spaink H. P., Wijffelman C. A. and Lugtenberg B. J. J., "A Rhizobium leguminosarum biovar trifolii locus not localized on the sym plasmid hinders effective nodulation on plants of the pea cross-inoculation group.," Mol. Plant-Microbe Interact., vol. 10, pp. 938-941, 1997.

[33] Russell A. B., Hood R. D., Bui N. K., LeRoux M., Vollmer W., Mougous and Joseph D., "Type VI secretion delivers bacteriolytic effectors to target cells," Nature, vol. 457, pp. 343-349, 2011.

[34] Sandra Schwarz, Rachel D. Hood and Joseph D. Mougous, "What is type VI secretion doing in all those bugs?," ELSEVIER, vol. 18, no. 12, pp. 531-537, 2010.

[35] Schwarz S., West T. E., Boyer F., Chiang W.-C., Carl M. A., Hood and et al., "Burkholderia type VI secretion systems have distinct roles in eukaryotic and bacterial cell interactions," PLoS Pathog , Published online, 2010.

[36] Silverman J. M., Agnello D. M., Zheng H., Andrews B. T., Li M., Catalano C. E. et al.,"Haemolysin co-regulated protein is an exported receptor and chaperone of type VI secretion substrates,"Mol. Cell, vol.51,pp. 584-593, 2013.

[37] Silverman JM, Brunet YR, Cascales E and Mougous JD., "Structure and regulation of the type VI secretion system.," Annu Rev Microbiol, vol. 66, pp. 453-472, 2012.

[38] Sophie S. Abby, Jean Cury, Julien Guglielmini, Bertrand Néron, Marie Touchon and Euardo P. C., “ Identification of protein secretion systems in bacterial genomes,” Nature, vol. 6, doi:10.1038/srep23080, 2016.

[39] Tao G. Dong, Brian T. Ho, Deborah R. Yoder-Himes and John J. Mekalanos, "Identification of T6SS-dependent effector and immunity proteins by Tn-seq in Vibrio cholerae," PNAS, vol. 110, no. 7, pp. 2623-2628, 2012.

[40] Thibault G. S., Kyler A. L. and Denise M. M., "T6SS: The bacterial "fight club" in the host gut," PLOS Pathogens, vol. 13, no. 6, 2017.

[41] Yi-Wei C., Lee A. R., Davi R. O. and & Grant J. J., "In vivo structures of an intact type VI secretion system revealed by electron cryotomography," EMBO report, Published online, 2017.

[42] William S. G., Varcoe L.. T., Attridge S. R., and Manning P. A. , “Vibrio cholerae Hcp, a secreted protein coregulated with HlyA.,” Infect. Immun, vol. 64, pp:238-689, 1996.

[43] Zheng J. and Leung K. Y., "Dissection of a type VI secretion system in Edwardsiella tarda.," Mol. Microbiol, vol. 66, pp. 1192-1206, 2007.

[44] Zoued A., YR Brunet, E. Durand, MS Aschtgen, L. Logger, B. Douzi and et al., "Architecture and assembly of the Type VI secretion system.," Biochim Biophys Acta, vol. 1843, no. 8, pp. 1664-1673, 2014.


13

IN VITRO AND IN VIVO STUDY OF BI-LAYERED LIDOCAINE AND EPINEPHRINE-ELUTING PLGA/COLLAGEN NANOFIBROUS

MEMBRANE

1FU-YING LEE, 2TZU-CHIA LEE, 3SHIH-JUNG LIU

1Department of Periodontics, Division of Dentistry, Chang Gung Memorial Hospital, Tao-Yuan 33305, Taiwan 2Department of Mechanical Engineering, Chang Gung University, Tao-Yuan 33302, Taiwan


Abstract- This study developed multi-layered lidocaine- and epinephrine-eluting biodegradable poly[(d,l)-lactide-co-glycolide] (PLGA)/collagen nanofibers. An electrospinning technique was employed to fabricate the multi-layer biodegradable drug-eluting nanofibers. After fabrication, the nanofibrous membranes were characterized. The drug release characteristics were also investigated. In addition, the in vivo efficacy of nanofibers for pain relief and hemostasis in palatal oral wounds of rabbits were evaluated. Histological examinations were also completed. The experimental results suggested that all nanofibers exhibited good biocompatibility and eluted effective levels of lidocaine and epinephrine at the initial stages of wound recovery. Index Terms - Biodegradable nanofibers; PLGA; collagen; epinephrine; lidocaine I. INTRODUCTION Free soft tissue grafts were widely used and highly predictable procedure in gingival augmentation procedure. But postoperative pain and excessive bleeding are the common complications following soft tissue grafting procedures [1-4]. The purpose of this study was to evaluate the efficacy of multi-layered lidocaine/epinephrine/collagen loaded biodegradable nanofibrous membranes for pain relief and hemostasis in palatal oral wound of rabbits. II. MATERIAL & METHOD 12 rabbits were randomly assigned to two groups. Free gingival graft was harvested from palatal mucosa. The wound was dressed by multi-layered biodegradable nanofibrous membranes [5] in test group and wound was exposed in control group. The following variables were measured: daily change in body weight, food and water intake before and after surgery. Histological examination at the wound site was also performed. III. RESULTS All membranes have good biocompatibility and eluted effective levels of lidocaine and epinephrine on the early-stage of wound healing [6-9]. Rabbits in the test group showed faster recover of food and water intake and hemostasis, post operatively compare with those in control group [10]. CONCLUSION Our multi-layered biodegradable nanofibrous membranes offer adequate efficacy and durability pain relief and hemostasis on the initial healing in rabbit

with palatal oral wound. REFERENCES

[1] Matter, J. Free gingival grafts for the treatment of gingival recession. A review of some techniques. J. Clin. Periodontol.1982, 9, 103–114.

[2] Keceli, H.G.; Aylikci, B.U.; Koseoglu, S.; Dolgun, A. Evaluation of palatal donor site haemostasis and wound healing after free gingival graft surgery. J. Clin.

Periodontol.2015, 42, 582–589. [3] Cairo, F.; Nieri, M.; Pagliaro, U. Efficacy of periodontal

plastic surgery procedures in the treatment of localized facial gingival recessions. A systematic review. J. Clin. Periodontol.2014, 41 (Suppl. 15), S44–S62.

[4] Roccuzzo, M.; Bunino, M.; Needleman, I.; Sanz, M.

Periodontal plastic surgery for treatment of localized gingival recessions: A systematic review. J. Clin. Periodontol.2002, 29 (Suppl. 3), 178–194; discussion 195–176.

[5] Chen, D.W.C.; Liao, J.Y.; Liu, S.J.; Chan, E.C. Novel biodegradable sandwich-structured nanofibrous drug-eluting membranes for repair of infected wounds: An in vitro and in

vivo study. Int. J. Nanomed.2012, 7, 763–771. [6] Malmquist, J.P.; Clemens, S.C.; Oien, H.J.; Wilson, S.L.

Hemostasis of oral surgery wounds with the HemCon Dental Dressing. J. Oral. Maxillofac. Surg.2008, 66, 1177–1183.

[7] Cunha-Reis, C.; TuzlaKoglu, K.; Baas, E.; Yang, Y.; El Haj, A.; Reis, R.L. Influence of porosity and fibre diameter on the

degradation of chitosan fibre-mesh scaffolds and cell adhesion. J. Mater. Sci. Mater. Med.2007, 18, 195–200.

[8] Shen, E.C.; Chou, T.C.; Gau, C.H.; Tu, H.P.; Chen, Y.T.; Fu, E. Releasing growth factors from activated human platelets after chitosan stimulation: A possible bio-material for platelet-rich plasma preparation. Clin. Oral Implant.

In Vitro and In Vivo Study of Bi-Layered Lidocaine and Epinephrine-Eluting PLGA/Collagen Nanofibrous Membrane


14

Res.2006, 17, 572–578. [9] Gentile, P.; Chiono, V.; Carmagnola, I.; Hatton, P.V. An

overview of poly(lactic-co-glycolic) acid (PLGA)-based

biomaterials for bone tissue engineering. Int. J. Mol. Sci.2014, 15, 3640–3659.

[10] Lee, F.Y.; Lee, D.; Lee, T.C.; Chen, J.K.; Wu, R.C; Liu, K.C.; Liu, S.J. Fabrication of Multi-Layered Lidocaine and Epinephrine-Eluting PLGA/Collagen Nanofibers: In Vitro

and In Vivo Study, Polymers 2017, 9, 416.


15

ASSESSMENT OF WEBSITE CONTENT BASED ON WEBOMETRICS RANKING: CASE STUDY OF THAILAND UNIVERSITY

1TAWATCHAI SUKSIDA, 2LALITA SANTIWORARAK

1Faculty of Communication Arts, 2College of Creative Design and Entertainment Technology

1,2Dhurakij Pundit University, Prachachuen Road, Bangkok, Thailand E-mail: [email protected], [email protected]

Abstract - Website is tool which presented information of organization. The most important thing of website is content due to content will communicate with user and explain everything which organization presented. In addition, Educational Institution also use website to presenting their information to outside too and there are many organization ranks website of education for several purpose Webometrics Ranking of World Universities is the ranking of education website which assessed accessibility of website content. In this paper, we study about website’s content of top 10 and bottom 10 universities in Thailand ranking and Webometrics visibility indicators. This article aims to evaluate the average times which user use in website, total of non-unique visitors of website and percentage of bounce rate of Thailand University’s website which ranked in Webometrics. The result should help website’s administrator manage their content and It may help them to move up their site in Webometrics Ranking of World Universities too. Index Terms - Webometrics, Website Visibility, Search Engine Optimization, SEO, Content. I. INTRODUCTION Today, people can search information which they want to know by using search engine tools easily and quickly, especially website information. It is a representative of an organization to present academic news, courses, student life and other useful information. Webometrics ranking of world universities is an organization which rank educational institution around the world by Consejo Superior de Investigaciones Científicas (CSIC), a research group in Cybermetrics Lab of Spain in 2004 [1]. Webometrics ranking of world universities rank educational institution 2 times per year, the first times is the end of January and another one is the end of July. The main objective of the ranking is supporting the open access concept and transfer knowledge from educational institution to whole society. Webometrics using 2 indicators to prove website information accessibility [2]. The first indicator is 50% of website visibility by evaluation from website impact score which the website impact is number of backlinks to the educational institution website. Webometrics use 2 tools for evaluate backlinks include Ahrefs and Majestic SEO and the other one is 50% of website activity. The activities of website include 10% of Presence, 5% of Openness and the last one is 35% of Excellent. II. LITERATURE REVIEW R.K. Pandey [3] selected 1 to 19 universities from Webometrics ranking of world universities of July 2010, Academic Ranking of World Universities (ARWU), Times Higher Education (TIMES) and 4 International Colleges and Universities (4ICU) to

comparing each other and the result is 35% of indicators of Webometrics can manage by website owner and 65% is evaluated from website visibility and google scholar. The evaluation of website visibility is number of backlink to website. The number of backlinks to website is a business which huge income in website searching business. In addition, Ahmad Bakeri, Abu Baka and Nur Leyni [4] study about website visibility of Webometrics by compare the backlinks between 1 to 30 of webometrics ranking and 30 from bottom in webometrics word ranking universities 2013 and the result is number of backlinks and number of website accessibility of top 30 in webometrics ranking more than 30 bottom in webometrics ranking. III. METHODOLOGY The methodology of webometrics ranking of world universities have two indicators to rank educational institution website and the one from two of webometrics indicators is visibility which evaluate by impact values. Impact values can get from Majestic SEO and Ahrefs database by get the total numbers of backlinks from outside educational institution website and bring them to normalize. After that, the maximum value between the two sources is selected. It’s present in Impact Rank on webometrics ranking table. SimilarWeb Tools is the website analyze tool. In this paper focus on total visits pages, average visit duration time and bounce rate of the website. The process of this article are, at first, get Thailand ranking from Webometrics ranking of world universities in January 2018 focus on visibility indicator of Thailand universities. Then, divide the data for 2 groups, first group is top 10 of Thailand

Assessment of Website Content Based on Webometrics Ranking: Case Study of Thailand University


16

universities rank and the last one is bottom 10 of Thailand universities rank. After that, sorting of impact ranking of website. Then, evaluate the data

from of top 10 and bottom 10 of Thailand universities and finding the relationship of them.

IV. FINDING AND DISCUSSION Top 10 and Bottom 10 of Thailand raking in Webometrics is shown on Table 1.

Thailand Webometrics

Ranking (Jan 2018)

Impack Rank in

Webometrics University University Website

1 634 Mahidol University www.mahidol.ac.th 2 654 Chulalongkorn University www.chula.ac.th 3 537 Chiang Mai University www.cmu.ac.th 4 432 Kasetsart University www.ku.ac.th 5 549 Khon Kaen University www.kku.ac.th 6 1268 Prince of Songkla University www.psu.ac.th 7 843 Suranaree University of Technology www.sut.ac.th 8 1590 King Mongkut's University of Technology Thonburi www.kmutt.ac.th 9 1164 Thammasat University www.tu.ac.th 10 828 Naresuan University www.nu.ac.th 168 22043 Sae Institute Bangkok www.bangkok.sae.edu 169 22847 Pattani Community College www.pncc.ac.th 170 21951 Asian University www.asianu.ac.th 171 23640 Siam Technology College www.siamtechno.ac.th 172 23640 Buriram Community College www.brcc.ac.th 173 23893 Maehongson Community College www.mcc.ac.th 174 24742 Songkhla Community College www.sk-cc.ac.th 175 24384 Yasothon Community College www.yasocc.ac.th 176 24005 Phanomwan College www.phanomwan.ac.th 177 25221 Pitchayabundit College http://pcbu.ac.th

Table 1. Top ten and bottom ten of Thailand university website in Webometrics ranking From Table 1, it presents the impact ranking of Top 10 and bottom 10 of Thailand university which get from Webometrics ranking of world university, January 2018 edition. The sorting of impact ranking of Thailand University is shown on Table 2.

Thailand Webometrics Rank (Jan 2018)

Impack Rank in Webometrics University

4 432 Kasetsart University 3 537 Chiang Mai University 5 549 Khon Kaen University 1 634 Mahidol University 2 654 Chulalongkorn University 10 828 Naresuan University 7 843 Suranaree University of Technology 9 1164 Thammasat University 6 1268 Prince of Songkla University 8 1590 King Mongkut's University of Technology Thonburi



17

170 21951 Asian University 168 22043 Sae Institute Bangkok

169 22847 Pattani Community College

171 23640 Siam Technology College 172 23640 Buriram Community College 173 23893 Maehongson Community College 176 24005 Phanomwan College 175 24384 Yasothon Community College 174 24742 Songkhla Community College 177 25221 Pitchayabundit College

Table 2. Top ten and bottom ten of Thailand university website in Webometrics ranking order by impack ranking The best impact ranking of Thailand are ranked in top 10 of Thailand university and impact ranking of bottom 10 are put on the last table of ranking. From Table 2, the first ranking of impact rank of Thailand university are set on top 10 of Webometrics ranking, Kasetsart University is the first of the impact ranking table and the last ranking of impact rank of Thailand university are set on bottom 10 of Webometrics ranking too. Table 3 present the data from SimilarWeb tool and the data are average duration times and bounce rate of university website.

Thailand Webometri

cs Rank (Jan 2018)

Impack Rank in

Webometrics

University

Similarweb tool Avg. Visit

Duration

Total Visits

Bounce Rate

1 634 Mahidol University 0:02:39 16,430,000

62.05%

2 654 Chulalongkorn University 0:04:05 10,390,000

52.89%

3 537 Chiang Mai University 0:03:57 8,657,000 55.55%

4 432 Kasetsart University 0:05:34 17,850,000

43.11%

5 549 Khon Kaen University 0:03:04 9,660,000 38.02%

6 1268 Prince of Songkla University 0:03:47 10,290,000

51.03%

7 843 Suranaree University of Technology 0:04:26 4,854,000 30.79%

8 1590 King Mongkut's University of Technology Thonburi 0:03:19 3,601,000 53.76

%

9 1164 Thammasat University 0:03:28 7,036,000 40.29%

10 828 Naresuan University 0:03:19 3,554,000 42.54%

168 22043 Sae Institute Bangkok 0:02:05 34,930 4.64%

169 22847 Pattani Community College 0:00:11 66,038 39.91%

170 21951 Asian University 0:01:39 < 5,000 24.56%

171 23640 Siam Technology College 0:00:14 135,390 37.78%

172 23640 Buriram Community College 0:00:00 133,450 67.78%

173 23893 Maehongson Community College 0:00:10 < 5,000 78.44%



18

174 24742 Songkhla Community College 0:00:14 16,083 25.42%

175 24384 Yasothon Community College 0:00:17 < 5,000 44.07%

176 24005 Phanomwan College 0:00:00 12,152 27.70%

177 25221 Pitchayabundit College 0:01:44 45,650 7.45% Table 3. Result of top ten and buttom ten of Thailand university website which get from Similar Web Tool

Table 3 present traffic analytics of university’s website include total visits pages, average visit duration time and percentage of bounce rate on website. Total visits pages is the total number of web pages which user access on website. Visit duration time is the time which user spent on university’s website. Percentage of bounce rate is the ratio of duration time when user open the web page and close that page immediately. We found more average time which user focus on webpage of top 10 than bottom 10 in webometrics ranking Impact ranking of webometrics show on Table 1 and the meaning of that is, if the university ranked in the top of impact ranking, that university will ranked on top of Webometrics ranking too. The sorting of impact rank is shown on Table 2. Moreover, Table 3 shown the traffic which get from SimilarWeb tool. The data from SimilarWeb tool are total visits, average visit duration time and percentage of bounce rate. Average visit duration time means average of during time when user surf on university website. If the university website have more average visit duration time, it mean that website have many interesting content which attract user too. Total visits pages shown the total number of pages which user access on university website. The last one is percentage of bounce rate, it present the percentage of visitors who enter the site and then leave bounce rather than continuing on to view other pages within the same site. If educational website have high percentage

of bounce rate, they have to consider about their content and should find solution to solve them. CONCLUSION The relationship between impact ranking of the website in webometrics and result from SimilarWeb tool which we found include if website have more average visit duration time, it present the during time which user travel in the site that may help website administrator know about time which user use in the site and the percentage of bounce rate help website administrator know about interesting content of website, if university website have many interesting content, the percentage of bounce rate will reduce. REFERENCES [1] Isidro F. Aguillo, About Us. [Online], Available :

http://www.webometrics.info/en/Aboutus, Jan 24 2018. [2] Isidro F. Aguillo, Methodology. [Online], Available :

http://www.webometrics.info/en/Methodology, Jan 24 2018. [3] R.K. Pandey, “Empirical Validation of Webometrics based

Ranking of World Universities”, International Journal of Computer Science and Information Technologies, pp. 580 – 584, 2014.

[4] Ahmad Bakeri Abu Baka and Nur Leyni, N. P. J., “Webometric Study of World Class Universities Websites”, Qualitative and Quantitative Methods in Libraries (QQML) Special Issue Bibliometrics and Scientometrics, pp. 105 – 115., 2015.


19

THE STUDY OF POTENTIAL AND CAPACITY OF MEDICAL TOURISM EMPLOYEES ON THAI TRADITIONAL MASSAGE AND SPA BUSINESSES IN THE AREA OF BANG LAMPHU, BANGKOK

NARIN YUENTHON

Hotel and Lodging Business Management, Faculty of Humanities and Social Sciences, Suan Sunandha Rajabhat University


Abstract: This research is to study the potential and capacity of medical tourism employees on Thai traditional massage and spa businesses in the area of Banglumphu, Bangkok. It is a quantitative research using questionnaires as a research equipment. The data is surveyed by 165 stakeholders from owners and employees on Thai traditional massage and spa businesses in area of Banglumphu, Bangkok. The results is found that the potential of health-oriented tourism, and Thai traditional massage and spa categorized by backgrounds, economics, and society associated with experience and income, but also the connection between organization and its management truly have an impact upon the development of employees. Therefore, In a way, the analysis on the declining potential ofmedical tourism, and Thai traditional massage and spa and the yearn to improve service mind skill , in terms of knowledge, A more knowledgeable an employee is, the more desire an employee willing to develop his/her service-minded skills. Not only the erudite ones, but those who lack skills, character, and experience aspire to be better at service skill as well. Keywords - Potential and capacity, Medical Tourism, Health Tourism, Thai Traditional Massage, Spa business I. INTRODUCTION Tourism is an industry that changes in accordance with social and economic dynamics as well as new emerging technologies. These factors contribute to the development of tourism to be up to date and up to the current world situations. Health care business is a new form of tourism and is interested by many countries. In 2015, the starting of ASEAN Community has led to a strong and growing trend in competition in the field of tourism and health care businesses in the ASEAN region, especially the expansion of investment into the country that focused on alternative health services, for example, Thai traditional massage and spa business. Looking back at 2013, the revenue of spa businesses in Thailand was estimated to be over 979.1 million U.S. dollars in which over 90% of them came from foreigners. However, in 2014, the revenue has been decreased due to the political situations and expected to be increase again more than 90% over the next years. Nevertheless, the export value of Thai’s spa businesses that the foreigners used the services during the year 2013 – 2015 have a tendency to increase much more. (Department of International Trade Promotion, Ministry of Commerce, Thailand: 2015) Although Spa and Thai traditional massage in Thailand is increasing due to the advantage of better resources and services compared to other countries in Asia, human resource is one of the most important factors in this field of business. At the same time, it is the weak point of this business as well, in terms of lack of service providers and proficient skills. According to The Center for Economic and Business Forecasting, there is a shortage of human resources and skilled employees in the field of spa business

because most of the skilled employees change their jobs frequently. Some businesses have to hire foreigners to do the work for them in which it will have an effect on the image and the quality of Thai spa and Thai traditional massage in the future. (Sunhetra, Chanthaburi, 2016) For this reason, it must be acknowledged that the human resources of the spa and massage business group are considered to be one variable that is related to the strength of this business. (Phahol Sakkatat,et al.) Bang Lamphu is a diverse area of tourist attractions and services, for instance, Khaosan Road which is famous and full of spa and Thai traditional massage places. Therefore, from the importance of the problems stated in the previous paragraph, the researcher is interested in studying the potential and capacity that the service providers have to have on spa and massage businesses in the area of Bang Lamphu in order to develop both capacity and potential of the employees in order to achieve the level of standard of service. II. DETAILS EXPERIMENTAL 2.1. Objectives 1. To study and experience the capacity and

potential of medical tourism employees on Thai traditional massage and spa classified by socio-economic backgrounds, organizations, and ways of managements of the directors.

2. To analyze the regression between the potential/capacity of the employees and the demand for skill development services.

2.2. Literature Reviews The core of literature review in this research is that the content is being described in detail by explaining

The Study of Potential and Capacity of Medical Tourism Employees on Thai Traditional Massage and SPA Businesses in the Area of Bang Lamphu, Bangkok


20

the concepts and academic principles in a systematic manner as follows. Tourism Authority of Thailand (2009) has classified Medical Tourism into 2 types in accordance with the purpose of travel. First of all, Health Promotion Tourism is Natural or Cultural Tourism by managing the time from the tour to do health promotion activities, such as, body massage, body steam, herbal compress, aromatherapy, hydrotherapy, mineral bath, etc. Another type of Medical Tourism is Health Healing Tourism which is similar to Health Promotion Tourism except for the activity that is focused more on treatment, medical treatment, and rehabilitation, for instance, physical examination, dental treatment, or even cosmetic surgery. Furthermore, Wanna Wongwanich (2003) mentioned more about the essential of the service includes location, facility, security, appropriate management system, the experience that tourists will receive after the service, preparedness of the medical team, as well as skilled staff. The quality of service needs to be emphasized as the essential idea concerning serviced staffs in which various scholars have proposed this concept.Gronroos has mentioned 6 criteria that can be used to evaluate the quality of the services as followed: 1. The sophistication of the employees 2. Attitudes and demeanors of the employees when coping with the problems 3. The convenience and flexibility of the place and time of the service 4.Trustworthiness whereby the service is satisfactory 5. The ability to face and cope with problems 6. Good reputation will be gained from customers trust through the good service in which it is consistent the 5 factors of the SERVQUAL evaluation. First factor is the substantial of the service, second is trustworthiness, third is response towards customers, fourth is the reliability, and lastly understanding the customers. (Chatchawan Orawongsuphatat, 2011) According to the concepts in the previous paragraph, they can also be used as a guideline in various fields of business as well as the academic approach. Based on the importance of the quality of service, especially in term of the service employee, Sunethra Chanthraburi (2016) researched about the opportunity and competitiveness of spa and Thai traditional massage industry based on the framework concept of Diamond Model. The research illustrates that the competitiveness of Thai businesses, in terms of production factors, a pleasant personality shown by the employee, expertise, and like to service, however, there is still a shortage of skilled employee and the problem of frequent job changes, despite, the human factor remains the main factor in the analysis of competition evidently. Likewise, Phahol Sakkatat,et al., also researched about the strength of Thai spa and Thai traditional massage businesses in Chiang Mai community by using the independent variables that are related to the strength of the Thai spa and traditional massage business. The research shows that

the human factors consisting of knowledge of the members and number of members with workload were at a high level, the mean was 3.80. Therefore, this indicates the importance of the employee as a driving force of the business which is in consistent with the research of Paleerat Karndee, e al., (2004) about the development of Medical Tourism (spa) that comes from the opinions from 4 groups of people including public and private sectors executives, tour group, spa business group, and spa tourists group. In terms of human factors, the research shows that serviced employee should be trained and tested up to the standards as well as being expertise and able to work efficiently and able to communicate in English and at the same time serviced employee should always develop new knowledge and technology frequently. 2.3. Methodology This research is a development research focusing on the sets of data that would lead to the development of potential and capacity of the serviced employees of Thai spa and Thai traditional massage businesses in the area of Bang Lamphu, Bangkok. Accordingly, the quantitative research based on the survey from serviced employees working on spa and Thai traditional massage businesses in Bang Lamphu, Bangkok- 14 shops and 165 staff. The questionnaire was used for collect data. The questionnaire was developed by researching documents, texts, related research and verified by the experts. The questionnaire was divided into 3 sections as followed: 1. General information of the respondents 2. Information about the potential and capacity of Medical Tourism staff 3. Recommendations by using Interval Scale system using the deviation standard of Likert Scale, on a scale of 1 to 5: 5 = the most, 4 = very much, 3 = neutral, 2 = a little, and 1 = the least. In terms of the interpretive criteria for measuring the real value of potential/capacity of spa and Thai traditional massage serviced employees, the average points are set as follows: Average point between 1.00 – 1.49 signified as the least 1.50 – 2.49 signified as a little 2.50 – 3.49 signified as neutral 3.50 – 4.49 signified as very much 4.50 – 5.00 signified as the most In addition, the questionnaire was then sent to verify the accuracy and reliability at the level of 95 percent. After all, 165 sets of information are collected and then sent to analyze the statistical data with the program called SPSS by using the ways to calculate the statistics are as follows:

1. Descriptive statistic was used to explain the basic information of the group by percentage.

2. Chi-Square Test was used to test the relationship between the 2 variables and Regression Analysis was then used to analyze the regression of the variables.



21

III. RESULTS AND DISCUSSION According to the collective data from 165 people in the community on the relationship between the potential/capacity of Medical Tourism employees on Thai spa and Thai traditional massage classified by background of the population, economy and society, and the image of the business and the management system using Chi-Square Test. According to regression analysis, serviced employees of Medical Tourism should have 5 capacity which are knowledge, skill, capability, pleasant personality and attitude towards skill development, the results are shown as follows:

Variable

Potential & Capacity

Less Neutral Much X2

P-Value

Sex Male

Female

0 0

35 54

24 52

1.071

0.192

AGE 20-30 years 31-40 years 41-50 years

More than 50 years

11 36 27 15

9 25 25 27

1.370

0.713

Education Below

Bechelor Bechelor

More than bechelor

10 4

50 29 6

30 31 5

1.193

0.176

Work experience 1-3 years 4-6years 7-10years

More than 10 years

2 11

25 38 19 6

24 14 16 10

3.497

0.010*

Income Below 500

Us. 500 – 1,166

Us. 1,167-1,833

Us. 1,834 Us.

4 2

15 42 10 4

11 55 18 2

3.166

0.024*

*Level of Significance at 0.05 Table 1: Illustrates the results of the potential/ capacity of employees working in medical tourism, Thai traditional massage and spa classified by backgrounds, economics, and society. Regarding the statistical data X2 from the abovementioned chart, it has been found that the potential/capacity of an employee in the health-oriented tourism, Thai traditional massage and spa is merely experienced in their respective career and the significance of income is stood at only 0.05.

Variable Potential & Capacity Less Neutral Much X2 P-Value

Size of Business Small

Middle Large

0 23 8

24 47 19

10 17 16

64.066

0.040*

Duration of Business

1 – 3 years 4 – 6 years

7 – 10 years More than 10

years

2 9 6 5

11 21 8 8

22 7 59 7

152.87

0.183

Customer/day 21 -30 persons 31- 40 persons More than 40

10 4

29 29 17

13 31 26

160.48

0.301

No. of Employees

1 – 5 persons 6-10 persons

11 - 15 persons More than 15

2 9 4

2 25 23 3

9 22 52 13

60.398

0.001*

Potential Development Promotion for

employee -No support

-Support employees to attend training

-Support employees to

test skill competency

standard -Have a

evaluation system

12 2 1 2

38 4 2 4

62 25

4

9

58.644

0.000**

Administration - Place and

environment - Facilities

- Conservation

12 6 24

18

36 17

15

28 9

60.426

0.031*

*Level of Significance at 0.05 Table 2: Showcases the results of potential/ capacity of employees in the field of medical tourism, Thai traditional massage and spa categorized by their organization and the management of its manager. From the statistical data X2 in the 2nd chart at the significance level of 0.05, the mentioned group above values business scale and management at 0.05. Potential/Capacity

of Employees Coefficient of Equation (B) Sig R2

Knowledge 0.350 0.000** 0.024 Skill -0.398 0.000**

Ability -0.147 0.080 Personality -0.280 0.000**

Attitude 0.163 0.064 *Level of Significance at 0.01 Table 3: analyses the deterioration of employees’ potential/ capacity and the determination to enhance their service mind skill. Of all the 5 factors in the 3rd chart, it has been discovered that only the aspect of knowledge is going in the same direction as the desire to aggrandize



22

service mind capability in the low significance level at 0.01 which the sig value is less than 0.05. In other words, the more an employee knows, the more zealous to be more professionally competitive an employee will be. At the same time, the aspects of skills and character are the antithesis in the low significance level at 0.001 (sig = 0.000) which the sig value is less than 0.05 meaning the desire to better service mind skill will not be prominent in an employee with unpleasant personality, exposed to a small amount of work. The latter explanation can be expressed in the following equation of y = 1.287+0.350X1+0.398X2+0.280X3; R2 = 0.024 which can estimate the potential/ capability of an employee with the wish to improve one’s service mind skill at 2.4 percent. Therefore, it can be concluded that not only the relationship between the potential of health-oriented tourism, and Thai traditional massage and spa categorized by backgrounds, economics, and society associated with experience and income, but also the connection between organization and its management truly have an impact upon the development of employees which is relevant to Benjaporn Yaemjamuang (2014) who studied the topic of The readiness of staff’s development in housekeeping department for 3 stars hotel at Samutsongkram province to reaches to ASEAN community, the result stated that there are 3 competency affected to staffs in housekeeping department as follows ; the first, core competency in business administrative conformed to law and next, general competency in English skill and the last, core competency of work among multi social environment.As a matter of fact, this finding coincides with Kamolluck Lasuwong’s (2524, p. 229-230) statement on the potential development skill that besides a complete physical body and mind or so called maturity, worldly and inherent knowledge are necessary. Especially those who possess more experience have the tendency to progress in a constructive manner than the inexperienced ones. In a way, the analysis on the declining potential ofmedical tourism, and Thai traditional massage and spa and the yearn to improve service mind skill , in terms of knowledge, A more knowledgeable an employee is, the more desire an employee willing to develop his/her service-minded skills. Not only the erudite ones, but those who lack skills, character, and experience aspire to be better at service skill as well. That being said, the results reflect the service quality which largely contributes to employees’ self-development which aligns with the criterion of Gronroos who explicates on the professionalism and the skills of facilitators including their attitude and behaviors. Certainly, Gronroos’ criteria is also in accord with the evaluation of SERVQUAL which underlines the tangible of service, credibility, customer relationship management, and understanding customers (Chatchawan Orawong, January 5, 2018). Hence, the standard evaluation of

service quality encompasses knowledge, skills, and character of facilitators. Analyzing from the results above, it was found that the primary focus of a facilitator is to widen one’s knowledge and expertise in his/her respective field which correlates with Sunatra Chantaburi’s research on the opportunity and ability to compete in spa and Thai traditional massage’s industry by using the concept of Diamond Model. The competition amongst this type of business requires personnel who are not only cultivated with a solid foundation of service mind but also experience as well as capability. What’s even more challenging is managing the same set of workers due to the high frequency of employees’ quit rates. It has been found by Pahol Sukkatud and his committee that the strong establishment of spa and Thai traditional massage business by the community in Chiang Mai owes much to the know-how ofeach member accompanied by the balance between the number of members and the amount of work which is essential to forging a robust business. Moreover, the research carried out a decade ago on development of medical tourism which collected opinions of 4 interrelated groups of people including governmental and private organizations’ administrators, tour guide and Thai traditional massage and spa owners, and tourists by Paleerut Karndee and her associates (2547) is a reminder of the past about developing facilitators’ potential. To be specific, the research underlines that a group of entrepreneurs should pass standardized trainings and examinations until they are prolific, experienced, fluent in English, and tech-savvy. SUGGESTION Further research pertaining to characteristics of successful Thai traditional masseuse and spa business owners should be conducted. REFERENCES [1] Benjaporn Yaemjamuang. 2014. The readiness of Staff’s

Development in Housekeeping Department for 3 Stars Hotel at Samutsongkram Province to Reaches to ASEAN Community. (Research Project, Suan Sunandha Rajabhat University).

[2] 2.Chachawal Orawongsupatad. 2016. Implement of Service Qualification Study. Retrieved January 05, 2018, from http://www.tpa.or.th/writer/read_this_book_topic.php?bookID=1285&pageid=7&read=true&count=true

[3] 3.Kamolrat Lhasuwong. 1981. Educational Psychology. Bangkok: Srinakharinwirot University (Prasanmit Campus).

[4] Phahol Sakkatat et al. 2012. Strenght of Community’s Spa and Thai Traditional massage Businerss in Chiang Mai. (Research Project, Maejo University).

[5] Paleerat Karndee et al. (2004). Health Tourism Research. Bangkok: Ramkamhaeng University.

[6] Sunatetra Vhantaburee. 2016. Opportunity and competitive Competency of spa and Thai Traditional Massage. 17(2), 56-60.

[7] Tourism Authority of Thailand. 2009. Tourism Authority of Thailand. Retrieved January 05, 2018, from http://tat.or.th.


23

THE STUDY ON ELDERLY TOURIST ON TOURISM ACTIVITIES IN RANONG PROVINCE

TEERA INTARARUANG

Faculty of Humanities and Social Sciences, Suan Sunandha Rajbhat University


Abstract - Nowadays, Elderly tourist is potential and interesting market due to society approach to elderly society. United Nation estimated that elderly people will increase from 687.9 million in 2549 B.E. to 1,968 million in 2593 B.E. which Asia will have most elderly people. More over elderly people in Thailand will increase upward trend as 1/3 are Japanese tourist. These are good opportunities for Thailand to draw this elderly group. Elderly tourist behavior differences from general tourist. They emphasize on quality service which reasonable. They select activities which suitable for their age and health. Therefore, tourism program as Slow Tourism, close to nature, non-pollution healthy food, comfortable for elderly vehicles and prepared for safety should be created [1]. Ranong is located in the south part of Thailand, on the west coast along the Andaman Sea. It is the least populated province. Eighty percent of its area is covered by forests, and 67 percent is mountainous [2]. Ranong is one of the province which tourist should not missed due to its famous of natural resources and historical story. Ranong now become famous tourist destination, both Thai and foreigner tourist continuously come to visit. There are many famous tourist attraction such as Phayam Island, NamtokNgao National Park or 2 natural hot springs; Hot Springs Raksawarin and Porn Lung Hot Springs [3]. In addition, Ranong province is intend to be “A hub of transport goods between countries Andaman coast countries and preservation of city of Health Tourism and be livable environment” [4] The province established large size of hospital “Ranong Hospital” which fully with facilities for patientand “Thai Traditional Medicine and Herbal product and demonstration Centre” for preserve and rehabilitate people health [4]. Therefore, Ranong province is another one destination that elderly tourist interested. Hence, researcher then would like to study on elderly tourist interested toward tourist activities when they traveling purpose to created new tourism program which satisfy elderly tourist group purpose to draw this potential group to come to visit Ranong Province. To accomplish the goals and objectives, qualitative research was applied. The study also used both primary data and secondary data. Research result founded Tourist activities which got the most mean scores (4.73) are infusion in hot springs indicated most interested. Secondly, body rejuvenation Class got mean scores as (4.69) considered most interested. Thirdly belong to take a boat trip and dressing local costume with mean scores closed to seconded place (4.67) indicated most interested. Then health’s check program got the mean score (4.58). Follow by visit beautiful water fall with mean score (4.53). Last top five are cooking healthy food class with mean scores (4.24) considered most interest. Then join local people to make local souvenir (4.23) represented most interested. Follow by cultural route day trip got mean scores as (4.18) signify very interest. Before last is join local people to make dessert as mean scores (4.05) carried very interested. The lowest score belong to meditation class with mean scores as (4.02). Keywords - Ranong Province, Health Tourism, Elderly Tourist, Hot Spring. Tourist Activities I. INTRODUCTION Ranong Province is located in south part of Thailand far from Bangkok about 568 kilometre. As above-mentioned, at present it is well- known as attractive tourist destination due to their tourism resources especially the 2 famous hot springs Raksawarin and Porn Rang hot springs which attract large number of tourists to visited. Lonely Planet, the famous travel magazine stated that “The Andaman's northernmost province is a whole different package to the white sand, turquoise sea paradise that is used to sell the Andaman's on tourist brochures [5].” In order that vision and mission of Ranong province that intend to be “A hub of transport goods between countries Andaman coast countries and preservation of city of Health Tourism and be livable environment” [6] The province is intended to promoted their province as city of health tourism by used their potential tourism resources and established “RanongHospitalo” and “Thai Traditional Medicine

and Herbal product and demonstration Centre” for preserve and rehabilitate people health[7].Thus, it could say that Ranong province has potential enough to be an elderly tourist destination.

Fig. 1 Ranong Hospital Source: Pornnapat Berndt, November 24, 2017

The Study on Elderly Tourist on Tourism Activities in Ranong Province


24

Fig. 2 Water Lane for patient to excercise, Ranong Hospital

Source: Pornnapa Berndtt, November 24, 2017

Fig. 3 Local people infusion in Porn Rung Hot springs

Source: Pornnapa Berndtt, November 24, 2017 International Institute for Trade and Development stated that “Nowadays, Elderly tourist is potential and interesting market due to society approach to elderly society. United Nation estimated that elderly people will increase from 687.9 million in 2549 B.E. to 1,968 million in 2593 B.E. which Asia will have most elderly people. More over elderly people in Thailand will increase upward trend as 1/3 are Japanese tourist. These are good opportunities for Thailand to draw this elderly group. Elderly tourist behavior differences from general tourist. They emphasize on quality service which reasonable. They select activities which suitable for their age and health. Therefore, tourism program as Slow Tourism, close to nature, non-pollution healthy food, comfortable for elderly vehicles and prepared for safety should be created” [8]. Thus, this research hence intended to study on elderly interested on tourist activities in Ranong province aim to created program tour which appropriated and satisfy elderly tourist requirement.

II. RESEARCH METHODOLOGY This research focused on elderly people interested toward tourist activities in Ranong Province. The research instruments used are observation, questionnaire and content analysis, and in-depth interviews of elderly people. The study also uses both primary and secondary data. The sample of this research was calculated by using Taro Yamane (Yamane, 1973) [9]formula with 95% confidence level. (Refer elderly person ages from 60 years old up from Bangkok by sampling method) After calculated the sample size by using Yamane formula, the numbers of sample is 400persons. As Sample group is homogeneity group therefore researcher used sample only one fourth of sample which are 100 elderly persons. Researcher did research study based on information collected from 100 questionnaires. A Statistical Package for Social Sciences (SPSS) was used to analyze the data. Descriptive and inferential statistics such as tables, percentage, mean and standard deviation were used for data analysis and summary. For Analysis and interpretation of data, the symbols and abbreviation will be utilized as follows: % =Percentage x ̅=Mean SD. =Standard Deviation The analysis and interpretation of data were presented in form of tables. The objectives of the study were as follows: PART 1: The analysis of general background of respondents PART 2: The analysis of “Elderly interested in tourist activities” by descriptive statistics which were mean and standard deviation. The respondents were requested to score each of requirement on accommodation and facilities on the levels of requirement. The interval of the five-point Likert-type scale was set from 1 to 5 (5 is the most interested, and 1 is lowest interested).Besides, the translation of level ranking was analyzed by pursuing criteria of elderly person toward requirement on tourist activitiesByBest[10]. The scores between 1.00-1.80 mean the lowest interested The scores between 1.81-2.61 mean low interested The scores between 2.62-3.41 mean average interested The scores between 3.42-4.21 mean very interested The scores between 4.22-5.00mean the most interested Part 1: The analysis of general information of 263 respondents by descriptive statistics, which were frequencies and percentage, showed in Table 1 as follows:



25

General Information Amount Person

Percentage (%)

Gender Male

Female

23

23

77 77 Total 100 100.00

Table 1showed general backgroundof100 Respondents Table 1 showed data analysis of general information of 100 respondents. There were 100 respondents, 23% male and 77% female.

Table 2 showed degree of elderly interested on Tourist Activities in

Ranong Province. From table 2, research result founded tourist activities which got the most mean scores (4.73) are infusion in hot springs indicated most interested. Secondly, body rejuvenation Class got mean scores as (4.69) considered most interested. Thirdly belong to take a boat trip and dressing local costume with mean scores closed to seconded place (4.67) indicated most interested. Then health’s check program got the mean score (4.58). Follow by visit beautiful water fall with mean score (4.53) Sixth is spa and massage got the mean scores as (4.44) considered most interested. Seventh, exercise for elderly class (4.35) then swimming at the beach (4.32). Next, make merit and give food to the monk (4.29). Followed

by Cooking Healthy Food Class got the mean scores as (4.24) respectively. All indicated as most interesting. Last top five are cooking healthy food class with mean scores (4.24) considered most interest. Then join local people to make local souvenir (4.23) represented most interested. Follow by cultural route day trip got mean scores as (4.18) signify very interest. Before last is join local people to make dessert as mean scores (4.05) carried very interested. The lowest score belong to meditation class with mean scores as (4.02). Research resulted shown that while elderly people concern to their health refer to most interested is to infusion in hot springs and body rejuvenation class, however they also interest take a boat trip and visit beautiful waterfall also. Refer to research resulted, It noticed that tourist activities that elderly have to joined with other people it quit got lower scores such as join local people to make dessert and souvenir or cooking healthy food class. Then it should create activities that their relative involved to make them more confidence. Health check program also got most interested which indicated that they prefer tourist activities which safe and good for their health. This resulted is conform to research result of Dr. Pornapat Berndt on research Title of “The Study on Elderly Tourist Requirement on Accommodation and Facility in Ranong Province” which stated that “most respondents concern in safety issues especially in bath room. They need hotel room’s physical with safety environment. In addition, healthy issue is important for them too as the most requirements on hot spring in their hotel.” [11].

Fig. 4 and5 Boat trip and dressing local costume

Source: Kemika Koompech, 2017 Hence, tourist activities should design for elderly by concern their physical and safety. Research resulted is useful for tour company for plan tour’s program that could satisfy them.



26

CONCLUSIONS AND DISCUSSION A. Conclusion As provincial vision and missionand tourism resources, Ranong Province is one of potential province which could draw tourist both Thai and foreigner to come to visit Tourist activities are one important component in tourism sectors, which good program, it could satisfy elderly tourist to re-visits which mean income generated in local communities. B. Discussion This research resulted was not cover in all contents. However, it could be reflect of elderly interest in tourist activities in Ranong province which useful for tour company to design activities which suit for elderly tourist group SUGGESTION All stakeholder such as Tourism Authority of Thailand (TAT), Education institution, Hopitality Sector, etc. should preparedness for support elderly tourist group in the future. Tourism Sector in Ranong province should do research and improve tourism attraction site and prepare physical for elderly group such as ramp, or handrail in walk way area include public toilets which easier to assess. ACKNOWLEDGMENT Researcher deeply appreciated to Mrs. Jeerawan Sewatadul, President of Ranong Tourism Association for full support for information. Next, really thank you to Ms. Thitima Nakbupha, Operation Thai Traditional Medical at Ranong Hospital for her kindness and valuable information. Next, Miss Kemika Koompechand her father for all support, I really appreciated. Researcher would like to thank you for research budget supported by Suan Sunandha Rajabhat University

Fig. 5 Miss Kemika Koompech, [Left]

Source: Pornnapat, November 23, 2017

REFERENCES [1] International Institute for Trade and Development. Elderly

Tourist Market: New Chance for Thai grow for support AEC [Online], accessed 5 December 2017. Available from http://www.itd.or.th

[2] Wikipedia, The Free Encyclopedia. Ranong Province. [Online], accessed 5 December 2017. Available from https://en.wikipedia.org/wiki/Ranong_Province

[3] 12 Cities which do not miss. Completely Enjoy. [4] Andaman seasides City[Online], accessed 5 December 2017.

Available from http://www.citieshiddengemsthailand.com [5] Governor's Office of Ranong. Vision and Mission [Online],

accessed 5 June 2017. Available from http://www.ranongcities.com/index.php?cmd=contentdesign&mode=vision

[6] Lonely Planet. Ranong Province [Online], accessed 29 November 2017. Available from https://www.lonelyplanet.com/thailand/andaman-coast/ranong-province

[7] Governor's Office of Ranong. Vision and Mission [Online], accessed 5 June 2017. Available from http://www.ranongcities.com/index.php?cmd=contentdesign&mode=vision

[8] Ms.ThitimaNakbupha, Operation Thai Traditional Medical at Ranong Hospital. Interview 24 November 2017.

[9] Governor's Office of Ranong. Vision and Mission [Online], accessed 5 June 2017. Available from http://www.ranongcities.com/index.php?cmd=contentdesign&mode=vision

[10] Yamane, Taro.1973.Statistics: An Introductory Analysis.Thirdeditio. Newyork : Harper and. Row Publication.

[11] Best, J. W., Research in Education, 3rd ed., (New Delhi, Prentice-Hall, 1977), 174.

[12] Pornnapat Berndt (Ph.d). The Study on Elderly Tourist Requirement on Accommodation and Facility in Ranong Province. 2017


27

SENIOR TOURIST’S NEEDS CLASSIFIED BY AGE, TOWARDS SPECIAL ACTIVITIES PROVIDED IN ACCOMMODATION IN

AMPHAWA, SAMUT SONGKHRAM

ONNAPAT MEUNPAOPONG

Hotel and Lodging Business Management, Faculty of Humanities and Social Sciences, Suan Sunandha Rajabhat University E-mail: [email protected]

Abstract - This research is to study Senior Tourist’s Needs classified by age. towards Special Activities Provided in Accommodation in Amphawa, Samut Songkhram. It is a quantitative research using questionnaires as research equipment. There are 400 key informants from tourists in Amphawa District, Samut Songkram. The results is found that tourist with 25-35 years of age has more opinion on needs of hotel service and hotel business development for elderly in special activities than tourist from 36-45 year-old group. And tourist with 36-45 years of age has less opinion on this factor less than tourist from 46-55 year-old group. Therefore, tourists who are more than 65 years old has more opinion on needs of hotel service and hotel business development for elderly in special activities than the tourists who are less 25 years of age, 36-45 years of age and 56-65 years of age. Keywords - Senior Tourist’s need, Special Activity, Accommodation, Amphawa I. INTRODUCTION Thailand is one of the countries that will become the aging society. It is predicted that, in 2020, the number of elderly population with more than 60 years of age will increase more than 13 millions. It is 19% from only 14% in 2013. Therefore, there has been readiness preparation in many aspects including tourism with management of many components to support elderly group which provides more convenient and safe and which is also the opportunity to new markets of tourism. Hotel and lodging business plays important role in tourism industries. There has been obvious changes and contributions in oder to support senior tourists in both facilities and other services. For example, Amari Watergate Hotel where accommodations are customized for elderly in accordance with standard set by Hotels Association. This group of tourists is considered having high purchasing power. So, they are one of interesting target to service industries. Apart from facilities that are core factors of readiness preparation for elderly, special activities provided in lodge that serve elderly are also as interesting and important as staying. Many places including Wanathara Resorts and Spa in Phitsanulok provide rooms and health activities which some among them are appropriate to elderly. This also becomes attractive to many tourists. SamutSongkhram has an increasing rate of tourists. It is not so far from Bangkok and also surrounded by suitable scenery and nature as well as various types of accommodations. However, there has not yet been much readiness preparation of special activities. The researchers are aware of this potentiality and would like to study senior tourists’ needs towards special activities served in lodging business in Amphawa, Samut Songkhram.

II. DETAILS EXPERIMENTAL 2.1. Objectives 1. To study senior tourists’ needs of special activities categorized by age group, provided by lodging business in Amphawa, Samut Songkhram 2.2. Literature Reviews The research about guidelines for long-stay tourism promotion and development held by Thammasat University Research and Consultancy (2002) indicates factors related to decision making in traveling and long staying in Thailand. Other opinions towards Japanese tourists’ needs in regards to factors influencing rise of long staying are activities done during stay in foreign countries to prevent loneliness and homesick. The activities that Japanese are interested in are: volunteering in several careers, experiencing lifestyle of local people, learning local cultures and traditions, traveling, relaxing, having health recovery and adventures like trekking. The research of Horneman et al. (2002 cited in Suan Sunandha Rajabhat University, 2008) found that the tourists with more than 60 years of age who lived in Queensland have various needs. Nevertheless, the mutual need is package tours including many places. Those places are serene but attractive nature and countryside. Factors influencing travelers’ decision to travel are their wellness, desires to spend time with family and friends. It is interesting that these travelers do not concern about traveling expenses. Department of Tourism is an organization related to tourism industries that sets service standard provided to executive tourists, elderly, family with children, and pregnant women (2012). Department of Tourism (2012) gives a definition and sets more specific standard of accommodations for long-stay travel.

Senior Tourist’s Needs classified by age, towards Special Activities Provided in Accommodation in Amphawa, Samut Songkhram


28

The definition of accommodations for long-stay travel is “commercial establishment providing temporary accommodations for long-stay tourists including bed and board and other services related to stay and travel during one place to another. There are many types of this establishment such as hotel, resort, apartment, and guesthouse. The criterion can be described as follows. The 5th element, additional services have 6 criteria: - Common laundry room, consider location, interior design, area management, as well as electric devices such as washing machine, clothes dryers, etc. - Baggage service, consider space management and conditions, dimensions, and service systems - Book corner or reading room, consider room management or newspaper area - Additional services, there are common spaces and instruments both indoor and outdoor such as fitness center, sauna, bicycles, massage and spa room, etc. - Special activities based on occasion and convenience like health activities, cultural exchange activities - Connectivity with external agencies/ organizations. There are sufficient and various tourism business networks which are attractive and trustworthy. Internationally, there has been set of standard for tourism industries in ASEAN called ASEAN Tourism Standard. The objectives are to encourage ASEAN nations to have mutual service standard. This standard is set according to agreement of 10 ASEAN nations (Ministry of Tourism and Sports of Thailand, 2007). Details are: ASEAN Home Stay Standard in which home stay means alternative accommodations for family with home-like decoration. It gives opportunity for customers to experience family life, community lifestyles, as well as attractive tourist destinations. The standard is set with following 7 points. But 1 of 7 point said about special activity as follows: - the special activities that are provided by host of home stay for tourists must be eco-friendly, non-negative impacts and must be local or cultural related. Chanchit Thiansiri et al. (2012) studied potentiality of Slow Tourism market for senior tourists traveling in upper Northern Thailand and found that Thai and foreign tourists are of 60-64 years old. Their traveling results from bonus and pension. This age group does not have health issue that corresponds with criterion and definition of elderly as previously mentioned. Apart from demographic characteristics, it is found that issues towards seniors’ needs of tourism products, especially, issues related to development in support of elderly group are development of service personnel, physical improvement of tourism destinations, and development of other services to

facilitate elderly which includes safety. All of these are essential to support senior tourists and must be taken into account. So, it is obvious that every service business development does not have much difference in factors related to development. Piyachat Thonphaeng (2012), referring patterns of senior tourism business development in the case study of elderly in Central Thailand, found that there are 8 factors of such development which are service provider or tour guide, readiness and facilities of tourism destinations, service rendered, activities, entertainment, safety and transportation, and reliability of marketing management. 2.3. Methodology This is quantitative research with sample group of senior tourists who are 60-69 years old in Amphawa and nearby area. Sample group is categorized by Sampling Techniques in the case of unknown numbers but has large size of population. Scaling of 45% at 95% Confidence Level results to 400 people of sample group. Questionnaire is research tool. III. RESULTS AND DISCUSSION According to the collective data from 400 key informants in the topic of Needs in special activities provided in accommodation for supporting the elderly tourists, the results as follows

Demography Data. Number Perentage Gender

Male 192 48 Female 208 52

Age

Below 25 years 120 30

25 – 35 years 120 30

36-45 years 61 15.25

46-55 years 85 21.25

56-65 years 7 1.75

More than 65 years 7 1.75

Table 1: Number and percentage of gender and age data from 400 key informants

The above table shows the data of gender and age of key informants in Samutsongkram province, it found that the most elderly tourists are female (208 persons accounted for 52%) and male ( 192 persons accounted for 48%) and Age aspect found that the most tourists aged below 25 years ( 120 persons accounted for 30%) and average age between 25 -35 years (120 persons accounted for 30%) and the least is age between 56-65 years and also age more than 65 years ( 7 persons accounted for 1.75%)



29

Factors X S.D. Level of factors Rank

1. Need Hotel to provide health

activities within hotel i.e. Chi-

gong, Thai massage, spa

4.39 0.66 The highest 1

2. Need hotel to provide inter-

cultural activities. 4.27 0.64 The

highest 3

3. Need hotel to provide cultural study activities.


4. Need hotel to provide natural activities i.e. garden sight-

seeing.


Table 2: average and standard deviation of the opinion about hotel service needs on special activities.

From above table shows average and standard deviation of the opinion about hotel service needs on special activities. Overall is the highest common (x=4.23). Six aspects of highest ranking level are as follow: Need Hotel to provide health activities within hotel i.e. Chi-gong, Thai massage, spa (x=4.39), Need hotel to provide cultural study activities (x=4.30), Need hotel to provide inter-cultural activities (4.27), Need hotel to provide natural activities i.e. garden sight-seeing (4.23), knowledge skill i.e. cooking (4.14) and beauty activities (4.05).

Mor

e th

an

65 y

ears

4.71

-0.5

226*

(0

.020

)

-0.4

282

(0.0

56)

-0.5

825*

(0

.010

)

-0.3

322

(0.1

57)

-0.7

143*

(0

.020

)

-

56 -6

5 ye

ars

4.00

0.19

17

(0.3

91)

0.28

61

(0.2

00)

0.13

17

(0.5

57)

0.38

21

(0.1

04)

- -

46-5

5 ye

ars

4.38

-0.1

904

(0.0

67)

-0.0

960

(0.3

55)

-0.2

504*

(0

.018

)

- - -

36-4

5 ye

ars

4.13

0.05

99

(0.4

35)

0.15

44*

(0.0

45)

- - - -

25 -3

5 ye

ars

4.29

-0.0

944

(0.2

03)

- - - - -

Low

er

than

25

year

s 4.

19

- - - - - -

X

4.19

4.29

4.13

4.38

4.00

4.71

Age

Low

er th

an 2

5 ye

ars

25 -3

5 ye

ars

36-4

5 ye

ars

46-5

5 ye

ars

56 -6

5 ye

ars

Mor

e th

an 6

5 ye

ars

Table 3: Data showed needs of tourists on special activity classified by age.

From above table shows Data of needs of tourists on special activity classified by age, it found that the tourists who aged between 25-30 years and 36-45 years, have sig. = 0.045 which is lower .05 of significance level. It means that the tourists who aged between 25 -35 years, need hotel service in special activities differently from the tourists who aged between 36-45 years. Therefore, the tourists who aged between 25 -35 years, need hotel service in special activities more than the tourists who aged between 36-45 years The tourists who aged between 36-45 years and 46-55 years, have sig. = 0.045 which is lower .05 of significance level. Tourists who aged between 36 -45 years, need hotel service in special activities differently from the tourists who aged between 46-55 years. Therefore, the tourists who aged between 36 -45 years, need hotel service in special activities more than the tourists who aged between 46-55 years The tourists who aged more than 65 years with age lower than 25 years, age between 36 -45 years and age between 56 -65 years have sig. = 0.020, 0.010 and 0.020 which is lower .05 of significance level. It means that tourists who aged more than 65 years, need hotel service in special activities differently from the tourists who aged lower than 25 years, age between 36 -45 years and age between 56 -65 years. Therefore, the tourists who aged more than 65 years, need hotel service in special activities more than the tourists who aged who aged lower than 25 years, age between 36 -45 years and age between 56 -65 years. From the above research’s results, it comes to conclusion that tourists’ opinion towards special activities inside lodging can be categorized by following age groups. Tourist with 25-35 years of age has more opinion on needs of hotel service and hotel business development for elderly in special activities than tourist from 36-45 year-old group. The tourist with 36-45 years of age has less opinion on needs of hotel service and hotel business development for elderly in special activities less than tourist from 46-55 year-old group. The tourists who are more than 65 years old has more opinion on needs of hotel service and hotel business development for elderly in special activities than the tourists who are less 25 years of age, 36-45 years of age and 56-65 years of age. Tourists in Amphawa, SamutSongkhram, who have different level of education, have opinion on needs of hotel service and hotel business development for elderly in services from personnel, in environment and places and buildings, and in safety.The results



30

show that special activities are therefore ones of important components for hotel and lodging business which are the purpose of senior tourists which conformed to Kevin Wongleedee (2011) who studied the topic of The Study of Senior Foreign Tourists’ spending Behavior in Thailand. It stated that Most of senior tourists emphasized the health more than tourism especially in healthy food which is the factor that was relevant to the result. And This corresponds with the research’s study about guidelines for long-stay travel promotion and development by Thammasat University Research and Consultancy (2002) that mentioned other opinions on Japanese tourists’ needs influencing their decision on choosing long-stay accommodations. The activities such as volunteering, experiencing local lifestyle, rehabilitation, during stay in foreign country to prevent loneliness and home sick are ones of those needs. In accordance with the long-stay standard policy of the Department of Tourism (2009), the 5thcomponent which relates to additional services, has special activities on particular occasions among these services. The special activities like cultural exchange and health-related correspond with the above study result and comply with homestay standard set in eco-friendly, local and cultural-related special activities for tourists. In the same way, PiyachatThongpaeng (2011) mentioned the model of senior tourism development on the case study of elderly in Central Thailand and it is found that there are 8 factors and activity is one of the factor encouraging tourism. Therefore, the special activities are therefore marketing factor that deserves development as well as

other hotel and lodging business development. This corresponds with the research of ChanchitThiansiri et al. (2012) that studied about slow tourism’s marketing potentiality in senior tourist in Upper Northern Thailand and it is found that the development deserved in elderly support are service provider development and other-services development to facilitate tourists including safety. SUGGESTION There should be readiness of personnel to facilitate or administer special activities for elderly in hotel and resort. REFERENCES [1] ChanchitThiensiri. 2012. The potential evaluation of the

Slow Tourism Market for Elderly Tourists in Upper-North Region of Thailand.FeuAcedemic Review, 6(1), 49 – 62.

[2] Department of Tourism. 2009. Accommodation Standard for Longstay Tourism. Retrieved January 07, 2018 from http://www.tourism.go.th/assets/portals/3/files/LongstayStandard.pdf

[3] Kevin Wongleedee. 2011. The Study of Senior Foreign Tourists’ spending Behavior in Thailand. Research Project: National Research Council of Thailand.

[4] PiyachatThongpang. 2011. Model of tourism development for aging Case study of aging in central of Thailand. 1(2), 70-75.

[5] PrayoonBoonprasert. 2002. Guideline to Promote and Develop long Stay Tourism: Case Study of Elderly tourist who aged more than 50 years, traveled for rest. Bangkok: Thammasart University.


31

AN ANALYSIS OF MOBILE BANKING CUSTOMERS FOR A BANK STRATEGY AND POLICY PLANNING

BEHROOZ NOORI

1Department of Industrial Engineering West Tehran Branch, Islamic Azad University Tehran, Iran

E-mail: [email protected], [email protected]

Abstract-Online banking is increasingly common. Financial institutions deliver online services via various electronic channels, subsequently diminishing the importance of conventional branch networks. This study proposed an integrated data mining and customer behavior scoring model to manage existing mobile banking users in an Iranian bank. This segmentation model was developed to identify groups of customers based on transaction history, recency, frequency, monetary background. It classified mobile banking users into six groups. This study demonstrated that identifying customers by a behavioral scoring facilitates marketing strategy assignment. Then the bank can develop its marketing actions.Thus, the bank can attract more customers, maintain its customers, and keep high customers' satisfaction. Keywords- Data Mining;mobile data, mobile banking; customer segmentation I. INTRODUCTION The newly emerging channels of online banking and rapidly increasing penetration rates of mobile phones motivate this study(C. S. Chen, 2013). The internet has had a significant impact on financial institutions, allowing consumers to access many bank facilities 24 hours a day, while allowing banks to significantly cut their costs. Research has shown that online banking is the cheapest delivery channel for many banking services (Koenig-Lewis, Palmer, & Moll, 2010; Robinson, 2000). A number of studies have identified advantages to bank customers, including cost and time savings as well as spatial independence benefits (Koenig-Lewis et al., 2010). According to Gartner’s prediction of leading trends of 2012 in mobile applications, mobile commerce (m-commerce) remains the most important one. Gartner further forecasts that mobile devices will replace PCs as the main device to access the internet. As for the third quarter of 2012, IPSOS indicated that “The era of Multi-Screen has come, and smartphones account for the purchasing behavior of 65% of mobile device users.” According to that report, 66 percent of the smartphone holders in Taiwan access the internet via a smartphone at least once daily; approximately 57 percent of the customers perform mobile searches; and 40 percent of the customers shop via mobile phones(IPSOS, 2012). These statistics reflect vigorous growth in the scale of m-commerce. However, mobile banking remains in its infancy, and international adoption rates demonstrate the strong potential of m-commerce(FRB, 2012). Therefore, data mining for mobile banking is of priority concern for further developing mobile banking services (MBSs) (C. S. Chen, 2013). Moreover, recent developments in Internet connectivity have led to a renewed interest in Internet banking among specific groups of working individuals. Moreover, with the rapid development of mobile and smart phones, Internet banking has

become more conducive to many more individuals, since they can carry out their banking transactions anywhere and anytime (Govender & Sihlali, 2014; Lee & Chung, 2009). Mobilebanking, an extension of Internet banking, provides time independence, convenience, prompt response to customers and cost savings. These benefits serve as an opportunity for banks to increase consumer market through mobile services. Furthermore, mobile technologies, such as smart phones, PDAs, cell phones, and iPads have not only become ubiquitous, but also trendy among young adults(Govender & Sihlali, 2014). Moreover, in recent years the market orientation has changed to customer centric view. After realizing the importance of simultaneous use of various channels, banking and financial companies are now paying attention to mobile banking especially when it comes to maintenance of customer relationships (Sangle & Awasthi, 2011). The ability to identify customer’s most pressing need at a given moment of time is one of the promising propositions of mobile banking. Advanced mobile technologies help banks in offering new services like viewing account details, fund transfer, balance enquiry, loan details, bill payments, enquiry about credit card and demat account and add value to existing ones by disseminating the information at userdefined time and place(Sangle & Awasthi, 2011). Besides, banking was at the forefront of the service sectors that migrate customers from face-to-face transactions to computer-mediated transactions. With the development of m-commerce, similar expectations have been held out that much banking activity that is currently carried out online through fixed line internet terminals will migrate to mobile devices. The range of services that can be undertaken while mobile is likely to increase, and mobile phones are likely to evolve as ubiquitous payment devices(Koenig-Lewis et al., 2010; Wilcox, 2009). Market segmentation is one of the most important areas of knowledge-based marketing. In banks, it is

An Analysis of Mobile Banking Customers for a Bank Strategy and Policy Planning


32

really a challenging task, as data bases are large and multidimensional(Zakrzewska & Murlewski, 2005). Though a number of aspects have been studied for m-commerce, very little is reported regarding the customer segmentation of mobile banking from customer relationship management (CRM)perspective(Wong & Hsu, 2008). The knowledge of the key mobile user segments in financial sector is still lacking. This study attempts to add to the body ofliterature by data mining in mobile banking services(Sangle & Awasthi, 2011). In relation to customer-centric business intelligence, banks are usually concerned with the following commonMarketing and sales concerns(D. Chen, Sain, & Guo, 2012): Who are the most / least valuable customers to

the bank? What are the distinct characteristics of them?

Who are the most / least loyal customers, and how are they characterized?

What are customers’ transaction behavior patterns? Which services have customers purchased together often?Which types of mobile banking users are more likely to respond to a certain promotion mailing?

What are the sales patterns in terms of various perspectives such as services, regions and time (weekly, monthly, quarterly, yearly and seasonally), and so on? and

What are the user segments in terms of various perspectives(D. Chen et al., 2012)?

In order to address these marketing concerns, data mining techniques have been widely adopted,coupled with a set of well-known businessmetrics about customers’ profitability andvalues, for instance, the recency, frequencyand monetary (RFM) model, and thecustomer life value model(D. Chen et al., 2012).

In this article a case study of using datamining techniques in customer-centricbusiness intelligence for a bankwas presented. The main purpose of this analysis is to helpthe bank better understand its mobile banking customersand therefore conduct customer-centricmarketing more effectively. On the basisof anewsegmentation model, customers of the bank have been segmented into variousmeaningful groups. Accordingly, a set ofrecommendations was provided to thebank on customer-centric marketing (D. Chen et al., 2012).

II. LITERATURE SURVEY Banks operate in a competitive environment facing challenges in customer acquisition and service costs. In such an environment, the understanding and prediction of customer behavior in usage of services is becoming an important subject. The banks’ intention is to shift customers to technology enabled self-service channels like ATMs, internet banking

and more recently onto mobile banking services. Customers, these days are more and more pressed for time and they seek a channel that offers them convenience of anytime, anywhere banking and mobile banking services fits the bill very well. In Iran,mobile banking services seem to be high on priority for banks (Thakur, 2014). Particularlyin Iran, banking services on mobile banking were launched few years ago yet the usage of such services has not reached thedesired level.Therefore, it becomes more important to look for the customer segments. The studies conducted on bank information technology adoption render insufficient information about customer segmentation(Sangle & Awasthi, 2011). In this regard the current study tends to emphasize customer data mining framework and identify the mobile user segments. 2.1. Mobile Banking While the use of branch-based banking is still very popular, banks have other ways of providing customers with financial management services and one of them is mobile banking (Govender & Sihlali, 2014). The mobile phone as a channel for service consumption offers enormous potential since today, a mobile phone is an integral part of customers’ life and a growing number of these devices are also equipped with internet connection. Currently mobile banking services enable consumers, for example, to request their account balance and the latest transactions of their accounts, to transfer funds between accounts, to make buy and sell orders for the stock exchange and to receive portfolio and price information(Laukkanen, 2007).Hence it is necessary to investigate mobile banking customer segments. 2.2. Cross-sellingAnalysis The rationale for cross-selling, defined in the introduction as “the strategy of selling other products to a customer who has already purchased a product from the vendor” isnot only to “increase the customer’s reliance on the company and decrease thelikelihood of switching to another provider” but also to exert a generally positiveinfluence on the relationship with the customer, strengthening the link betweenprovider and user (Kamakura, Wedel, De Rosa, & Mazzon, 2003). Increasing product holding leads to anincreased number of connection points with customers, as well as increasing theswitching costs they would face if they decided to take their custom elsewhere.Increased product holding also creates a situation in which the company can get toknow it customers better through a greater understanding of buying patterns andpreferences. This, in turn, puts it in a better position to develop offerings that meetcustomer needs. Consequently, it is argued that cross-sellingincreases the total value of a customer over the lifetime of the



33

relationship(Ansell, Harrison, & Archibald, 2007; Kamakura et al., 2003). Cross-selling, and consequently cross-buying, is receiving considerable attention inboth research and management in the financial services industry. Denoting to termssuch as “bancassurance” and “allfinanz”, i.e. the sales of insurance products by banks, and on the other hand “assurfinance”, i.e. the salesof financial products by insurance companies, changes in the market such as deregulation and increasingcompetition have driven the once traditional financial service providers towardsincreasing provision of integrated financial services, that is, offering their customers aseamless service of banking, investment and insurance products(Mäenpää, 2012; Van den Berghe & Verweire, 2001). 2.3. Bank customer segmentation Market segmentation has become one of the most dominant concepts in both marketing theory and practice. In banking industry, like any other service industries, segmentation is considered as a major way of operationalizingthe marketing concept, and providing guidelines for a bank’s marketing(Edris, 1997). As theory, market segmentation is the process of dividing a market into distinct groups of individuals, or organizations, who share one or more similar responses to some elements of the marketing mix. The segmentation process calls for dividing the total market into homogeneous segments, selecting the target segments, and creating separate marketing programs to meet the needs and wants of these selected segments(Edris, 1997). The identification of segments allows the evaluation and refinement of a bank’s marketing strategy. The effectiveness of the segmentation process and strategy depends on identifying segments that are measurable, accessible, stable, substantial, and actionable(Edris, 1997). 2.4. CLV and RFM Analysis Customer segmentation is used in differentsettings, for instance, using customer segmentationfor estimating customer future valueas a part of customer lifetime value (CLV) in banking scope (Khobzi, Akhondzadeh-Noughabi, & Minaei-Bidgoli, 2014). Generally, customersegmentation is often used to obtain moredetails about different customers in bankingscope. Actually, according to these studiesdiverse groups of banks’ customers are identified by segmenting based on customers’ financial transactions(Khobzi et al., 2014). RFM analysis is a widely used method thatidentifies customer behavior and representscustomer behavior characteristics, and it standsfor the words: Recency, Frequency, and Monetary.Generally, these parameters are defined asfollows(Khobzi et al., 2014):

• Recency: The interval between the purchase and the time of analysis.

• Frequency: The number of purchases within a certain period.

• Monetary: The amount of money spent during a certain period.

These definitions are adaptable and can varyin different cases. In recent years, several researcherstried to extend the concept of RFManalysis, but there is lack of studies that analyze the customer segments and RFM analysis focusing banks over the mobile banking platform. Thus, although the increasing competitiveness in mobile banking is motivating an exponential growth in the number of studies, there is a call for studies that will help us to understand how customer behavior are formed in the mobile banking business in greater detail.

Moreover, the rapid development of data mining methodsenables using large data bases of customer data toextract the knowledge, supporting marketing decisionprocess. Asthe ability to acquire new customers and retain existing iscrucial, especially in the finance marketplace, thepossibility of customer segmentation by obtaining theinformation on unknown hidden patterns has a majorsignificance. Until now only few papers present usingof data mining techniques in banks. In ourwork, we consider application of a new RFM segmentation algorithmin this area(Zakrzewska & Murlewski, 2005). III. METHODOLOGY In this study, numbers of mobile banking users of a major bank in Iran were studied. These user demographics were shown in table I. Additionally, bank customer table was shown in table II.The proposed methodology utilizeda new segmentation methodology, as shown in Fig. 1. In this work, customer priority number (CPN) or RFMD as a new model of RFM, was introduced for first time. It is the



34

product of the recency (R), frequency (F), average transaction amount or monetary (M) and customer deposit (D) ratings: RFMD = R F MD The rationale ofthe proposed approach is that if customers have hadsimilar purchasing behavior, then

theyare very likely also to have similar RFMD values. RFMD values were used tocluster customers into groups with similar RFMD values. The scaling of R–F–M-D attributes was shown in table III.



35

RFMD refers to the customer current value. It calculated for each stored customer data (Table II). RFMD or CPN ranking was illustrated in table IV.The rankings given are normally scored on a scale

of 1-4. Therefore, CPN would be between 1 and 144. After the case priority number (CPN)was computed, customer current value could be determined.After RFMD computation, potential value of customer



36

based on future opportunitiesshould be estimated. The CPN and potential value of customer are main elements for customer segmentation (Fig. 2).

IV. CASE STUDY This workconsidered a bankcustomer records to conduct empirical research (Fig. 3). Three customers were selected to show methodology effectiveness. The real data of selected customers and related R, F, M, and Ds were shown in table V. RFMDs were computed and customer potential values were illustrated in table V. Meanwhile customer type and its marketing strategy were derived (Table VI). V. MANAGERIAL IMPLICATIONS The bank’s marketing and business manager, bank branch manager, or analysts can employ the segments to: Better understand customers. The bank can track

changes to customers’ life styles. Better customer knowledge and understanding are the cornerstones of effective and profitable customer management(Zuccaro & Savard, 2010).

Enhance the value of segmentation systems. Proactive segmentation systems are enhanced when they are updated regularly. This means that both demographic and transaction data are integrated into an ongoing process of customer segment management. Customer segments possess the built-in capacity to integrate demographic and transaction data. Up-to-date and relevant segmentation system provide the bank with invaluable data to plan new service offerings, predict customer reaction and determine profit levels on a segment-by-segment basis. Segmentation system enhances the bank’s capacity to employ customer knowledge in a more strategically effective manner(Zuccaro & Savard, 2010).

Improve marketing effectiveness. Without a sound segmentationsystem a bank would not be able to perform valid and reliable customerprospecting which in turn would seriously undermine the effectiveness andprofitability of customer targeting. The starting point for serious customerprospecting and targeting is the bank’s customer data and transaction database.It provides the analyst with invaluable behavioral information (use of mobile banking by each customer). In addition, the database will contain rudimentarysocio-

demographic data such as the customer’s age, sex, maritalstatus and some employment information. Customer prospecting and targetingcould be undertaken employing such data. Customers would be placed in groups. Many organizations have realized that byenhancing their customer database they can significantly improve their customer prospecting and increase the lift ofcustomer targeting strategies. Thus, segmentation is designed to exploit thepotential of the bank’s customer database. Once a specific customer segmentgenerated by RFMDsegmentation has been identified, it becomes relativelysimple to identify the customer prospects and target them with the appropriatestrategy and promotional tools(Zuccaro & Savard, 2010).

Develop effective communications. In the age of segmentation, developingan effective communication strategy is not a simple task. The nature and varietyof potential communication messages and media to transmit the messages hasgrown exponentially during the last two decades. In addition, mostorganizations, including banks, are abandoning traditional communicationmedia such as television and radio and opting for more specialized vehicles suchas the web. Segmentation provides the bank with a richer set of segments that can be described with animpressive level of detail. The refined segments along with detailed financial life style of its membersallow the bank to design tailor-made communication strategies(Zuccaro & Savard, 2010).

CONCLUSION Mobile phone handsets, which were initially used almost exclusively for voice calls are now often used to transmit data and undertake commercial transactions. In recent years, mobile phones have become very popular with a penetration rate in many of states of Iran. The term m-commerce has been widely used to describe a subset of e-commerce and refers to transactions with monetary value that are conducted via mobile devices(Koenig-Lewis et al., 2010). Iranian banks today face intense competition inside and outside Iran. This in turn has forced these banks to be more oriented towards their customers. The main focus of this study was on the customer segmentation. Banks which are marketing-oriented are notonly required to be aware of the needs of theircustomers, but they should be able to satisfyeffectively the needs of each identified customersegment. This study provides evidencethat segmentation of the customers is of great importance to banks inorder to identify the behavior of each segmentand provide certain marketing actions that best suitthese behaviors.The results of this study provide a practicalapproach to Iranian banks that



37

wouldhelp in determining the true segments of mobile banking customers(Edris, 1997). Furtheremore, one of the important factors for the success of a bank industry is to monitor their customers' behavior. The bank needs to know its customers' behavior to find interesting ones to attract more transactions which results in the growth of its income and assets. The RFM analysis is an approach for extracting behavior of customers and is a basis for marketing and CRM, but it is not aligned enough for banking context(Bizhani & Tarokh, 2011). So, this study introducednew RFM model toimprove understanding of bank customers. Furthermore, this paper presented a framework of segmentation by applying it to the customers of one of Iran’s major banks. Also, this paper presented a synthesized example of segmentation in the banking sector.The proposed model improved current understanding of mobile banking customers. Meanwhile, from a practical perspective, insights provided by the study can help mobile banking managersto managemobile users’ behavior. REFERENCES [1]. Ansell, J., Harrison, T., & Archibald, T. (2007). Identifying

cross-selling opportunities, using lifestyle segmentation and survival analysis. Marketing Intelligence & Planning, 25(4), 394-410.

[2]. Bizhani, M., & Tarokh, M. (2011). Behavioral rules of bank’s point-of-sale for segments description and scoring prediction. International Journal of Industrial Engineering Computations, 2(2), 337-350.

[3]. Chen, C.S. (2013). Perceived risk, usage frequency of mobile banking services. Managing Service Quality, 23(5), 410-436.

[4]. Chen, D., Sain, S.L., & Guo, K. (2012). Data mining for the online retail industry: A case study of RFM model-based customer segmentation using data mining. Journal of Database Marketing & Customer Strategy Management, 19(3), 197-208.

[5]. Edris, T. (1997). Services considered important to business customers and determinants of bank selection in Kuwait: a segmentation analysis. International Journal of Bank Marketing, 15(4), 126-133.

[6]. FRB. (2012). Current use of mobile banking and payments. www.federalreserve.gov/econresdata/mobile-devices/2012-current-use-mobile-banking-payments.htm.

[7]. Govender, I., & Sihlali, W. (2014). A Study of Mobile Banking Adoption among University Students Using an Extended TAM. Mediterranean Journal of Social Sciences, 5(7), 451.

[8]. IPSOS. (2012). Mobile internet and smartphone adoption statistics. available at: www. digitimes.com.tw/tw/dt/n/shwnws.asp?CnlID¼10andCat¼35andCat1¼andid¼310528.

[9]. Kamakura, W.A., Wedel, M., De Rosa, F., & Mazzon, J.A. (2003). Cross-selling through database marketing: a mixed data factor analyzer for data augmentation and prediction. International Journal of Research in marketing, 20(1), 45-65.

[10]. Khobzi, H., Akhondzadeh-Noughabi, E., & Minaei-Bidgoli, B. (2014). A New Application of RFM Clustering for Guild Segmentation to Mine the Pattern of Using Banks’e-Payment Services. Journal of Global Marketing, 27(3), 178-190.

[11]. Koenig-Lewis, N., Palmer, A., & Moll, A. (2010). Predicting young consumers' take up of mobile banking services. International Journal of Bank Marketing, 28(5), 410-432.

[12]. Laukkanen, T. (2007). Internet vs mobile banking: comparing customer value perceptions. Business Process Management Journal, 13(6), 788-797.

[13]. Lee, K., & Chung, N. (2009). Understanding factors affecting trust in and satisfaction with mobile banking in Korea: A modified DeLone and McLean’s model perspective. Interacting with computers, 21(5), 385-392.

[14]. Mäenpää, I. (2012). Drivers of cross-sectoral cross-buying behaviour among business customers. International Journal of Bank Marketing, 30(3), 193-217.

[15]. Robinson, T. (2000). Internet banking: still not a perfect marriage. Informationweek. com, 4(17), 104-106.

[16]. Sangle, P.S., & Awasthi, P. (2011). Consumer's expectations from mobile CRM services: a banking context. Business Process Management Journal, 17(6), 898-918.

[17]. Thakur, R. (2014). What keeps mobile banking customers loyal? International Journal of Bank Marketing, 32(7).

[18]. Van den Berghe, L., & Verweire, K. (2001). Convergence in the financial services industry. Geneva Papers on Risk and Insurance. Issues and Practice, 173-183.

[19]. Wilcox, H. (2009). Banking on the mobile. white paper, Juniper Research, Basingstoke, available at www. juniperresearch. com (accessed 23 October 2009).

[20]. Wong, Y.K., & Hsu, C.J. (2008). A confidence-based framework for business to consumer (B2C) mobile commerce adoption. Personal and Ubiquitous Computing, 12(1), 77-84.

[21]. Zakrzewska, D., & Murlewski, J. (2005). Clustering algorithms for bank customer segmentation. Paper presented at the Intelligent Systems Design and Applications, 2005. ISDA'05. Proceedings. 5th International Conference on.

[22]. Zuccaro, C., & Savard, M. (2010). Hybrid segmentation of internet banking users. International Journal of Bank Marketing, 28(6), 448-464.


38

ADVANTAGE OF MAKE-TO-STOCK STRATEGY BASED ON LINEAR MIXED-EFFECT MODEL

1YU-PIN LIAO, 2SHIN-KUAN CHIU

1Chairman’s Office, Winstar Display Corp., Taichung, Taiwan

2Ph.D. Program of Business, Feng Chia University, Taichung, Taiwan , 3Department of International Trade, Feng Chia University, Taichung, Taiwan

E-mail: [email protected], [email protected]

Abstract- In the past few decades, demand forecasting becomes relatively difficult because of the rapid changes of world economic environment. In this research, the make-to-stock (MTS) production strategy is applied as an illustration to explain that forecasting plays an essential role in business management. We also suggest that linear mixed-effect (LME) model could be used as a tool for prediction and against environment complexity. Data analysis is based on a real data of order quantity demand from an international display company operating in the industry field, and the company needs accurate demand forecasting before adopting MTS strategy. The forecasting result from LME model is compared to the common used approaches, times series model, exponential smoothing and linear model. The LME model has the smallest average prediction errors. Furthermore, multiple items in the data are regarded as a random effect in the LME model, so that the demands of items can be predicted simultaneously by using one LME model. However, the other approaches need to split the data into different item categories, and predict the item demand by establishing model for each item. This feature also demonstrates the practicability of the LME model in real business operation. Index Terms- Forecasting, linear mixed-effect model, make-to-stock, order demand, production strategy I. INTRODUCTION Demand forecasting is crucial for supply chain management. Production planning, inventory management, and manufacturing scheduling are typically formulated according to short- and long-term expected demand [1]. To reduce the occurrence of delivery delays caused by the “crowding out” effect of manufacturing processes, contemporary enterprises have gradually changed their production patterns from make-to-order (MTO) to make-to-stock (MTS), and increasingly fewer enterprises are using the MTO production strategy [ 2 , 3 ]. The MTO production involves commencing product production only after the customer places the order. The MTS production pattern entails a stocking-up production, in which a company manufactures products and stores them in inventory before customer orders are received. Subsequently, the company sells its stock as customer places orders. If a company receives orders requesting a high mix of products but in low volumes, it must be capable of forecasting their order demand accurately before attempting an MTS production strategy. Accordingly, the advantages of the MTS production strategy—including quick delivery, arranging a long-term manufacturing schedule, reducing the stock levels, and stabilizing product prices—can be realized. Worldwide, variation in customer demand has forced many manufacturers to adopt a high-mix low-volume production model. However, this type of enterprise is not as efficient as a low-mix high-volume enterprise. Therefore, determining how high-mix low-volume enterprises can enhance their business operation performance urgently requires a solution. Hence, accurately forecasting order demand is a fundamental to successfully applying the MTS

production strategy to a high-mix low-volume business operation model. Because inaccurate demand forecast is a concern for high-mix low-volume enterprises, the MTO production strategy is typically adopted. However, this production pattern increases financial risks and requires a long delivery time, making centralized production difficult, which subjects production lines to frequent changes, resulting in high operating costs and low product quality. Complex operations are the primary cause of human error and low job satisfaction. Therefore, if the inefficiency of the high-mix low-volume business operation model cannot be solved, then, despite a high business revenue, business operation costs would increase rapidly, product quality would reduce, and employee job satisfaction and customer satisfaction would decrease, which result in that business development would stagnate. Therefore, the forecasting method proposed in this study can provide a crucial basis for transitioning from using the MTO to the MTS production, and may offer a viable solution for improving the business operation performance of high-mix low-volume enterprises. The application and improvement of the proposed forecasting method can assist researchers with understanding the characteristics of business operations and construct related business operation models. Forecasting ability depends on crucial information and reliable forecasting methods. In recent years, demand forecasting has become increasingly complex, primarily because the global economic environment has gradually changed. The underlying reasons for this change can be explained in terms of the following four dimensions: volatility, uncertainty, complexity, and ambiguity (VUCA) [ 4 , 5 , 6 ], all of which have been shown to influence

Advantage of Make-to-Stock Strategy Based on Linear Mixed-Effect Model


39

demand forecasting [7]. Volatility means that new products are rapidly developed, product lifecycles are shortened, customer preferences change suddenly, and organizations are frequently restructured; consequently, historical data diminishes in value. Uncertainty refers to unknown factors that cause sudden shifts in demand, and these factors are generally regarded as outliers or interferences. Complexity means that the interaction of these influential factors cannot be modelled easily, and ambiguity refers to fuzzy events and situations that cannot be quantifiably defined, leading to the loss of key influential factors. In summary, according to the influence of economics on demand forecasting, developing a reliable forecasting method requires analyzing whether historical data can contribute to demand forecasting, and whether the effects of influential factors can be identified. To meet the requirements of modern forecasting methodologies, this study proposed using linear mixed-effect models to perform forecasting. Linear mixed-effect models have been extensively developed and widely applied in various fields. However, no study has used this model to forecasting in business operation. Linear mixed-effect models are characterized by the inclusion of temporal factors and explanatory variables and the analysis of their significance. Accordingly, crucial influential factors can be identified to forecast demand. These characteristics fulfill the requirements of modern forecasting methodologies and can be used as the basis for companies to improve their operation efficiency and to develop competitive advantages. The following sections explore the influences of the MTO and MTS production strategies on business operation as well as the role of forecasting in the MTS strategy, provides a review of the literature on forecasting methodologies, and summarizes the strengths and weaknesses of commonly used forecasting methods. In addition, the proposed linear mixed-effect model as well as a method for model parameter estimation are introduced. Subsequently, the order demand of a manufacturer in central Taiwan is forecasted using product type as a crucial explanatory variable. Specifically, the linear mixed-effect model is applied to forecast the order demand for 20 individual product types. A 1-year forecast of monthly demand is reported, and three types of forecast errors are used to assess the forecasting ability of the model. The results show that the forecasting ability of the linear mixed-effect model in an empirical analysis is superior to those of a linear forecasting model, exponential smoothing method, and time-series forecasting method. II. LITERATURE REVIEW A. Influences of the MTO and MTS on Business Operations Modern production strategies primarily involve two

main production patterns: the MTO (based on customer orders), and the MTS (based on production capacity) [8]. From the perspective of customers, one competitive advantage of the MTS production is short delivery time and quick response [9]. Therefore, identifying the types of products that are specifically suitable for the MTS production pattern or both MTS and MTO patterns is a favored research topic in management science [8]. Regarding the influences of the MTO and MTS production strategies on business operations, Hendry and Kingsman [10] showed that the MTS and MTO production strategies are mostly used for manufacturing standard and customized products, respectively. Regarding the attributes of orders, order demand for MTS products is generally predictable, whereas that for MTO products is irregular and unpredictable. Concerning production planning, MTS production lines operate according to forecast results, and the production line schedule can be adjusted easily. However, the schedule of MTO production lines is determined based on recent order demand, and long-term manufacturing schedules are difficult to determine. In terms of product delivery, enterprises that adopt the MTS production strategy can ensure rapid product delivery, thus maintaining high customer satisfaction. The MTO production pattern requires long delivery times, and enterprises adopting this strategy must communicate with customers to achieve consensus regarding product delivery time. Concerning product price, compared with prices of products produced adopting the MTO strategy, the prices of MTS-produced products are relatively more stable. Soman, van Donk, and Gaalman [8] indicated that the MTO production pattern is effective for handling orders requesting high-mix customized products; the production planning for the MTO strategy must prioritize meeting order demands, while production effectiveness is determined according to crucial elements in the orders (e.g., the expected delivery volume and number of delayed delivery days). The goal of a company that manufactures MTO products is to shorten product delivery times; production efficiency emphasizes the importance of capability planning, orders that are lost due to problems with manufacturing processes, and on-time product delivery. By contrast, the MTS production pattern is effective for handling uniform product specifications and less customized products, where production planning is determined based on product demand forecasting and production effectiveness is production-oriented. Therefore, the goal of a company manufacturing MTS products is to enhance product availability, and its production efficiency emphasizes the importance of inventory policy, finished goods inventory, one-off or batch production, and accurate demand forecast. Rajagopalan [11] indicated that inventory costs are slightly higher for the MTS strategy than for the MTO strategy, particularly for



40

one-off and batch production. In summary, the MTS strategy relies heavily on the accuracy of product demand forecasting. Because of accurate forecasting, the advantages of the MTS production strategy, including short delivery time, manageable long-term manufacturing schedule, and stable product prices, can be realized. In addition, accurate forecasting can optimize inventory levels; therefore, companies applying the MTS strategy can effectively control inventory costs. Some researchers have explored the inventory policies and material control mechanisms in MTO production [12]. The forecasting method proposed in this study provides a relatively accurate basis for forecasting random customer orders (demand) for MTS production. B. Forecasting Methodology Two main types of forecasting methodology exist: (1) statistical methods; and (2) data mining and machine learning [13]. Both types of forecasting methodology are aimed at identifying the relationship between influential factors (independent variables) and research variables (dependent variables), and identifying the effects of the influential factors on research variables [7]. These two methodologies involve distinct approaches to interpreting analysis models. The statistical methodology is based on the data derived from a specific mathematical model as well as unobservable errors. The machine-learning methodology avoids fitting data to a specific model and develops algorithms that are suitable for various types of data. These two methodologies differ in their strengths and characteristics [13]. The statistical methodology uses the probability distribution of errors to infer the significance of the influential factors in a model. The reliability of inferences correlates positively with the mathematical model. The machine learning methodology uses the size of forecast errors as a basis for selecting the optimal forecasting model. Several typical forecasting methods are introduced as follows, the characteristics of which are shown in Table 1. The exponential smoothing method was proposed by Holt [14] and the statistical theoretical foundation for this method was established by Muth [15]. This method involves using a demand observation and predictive value in the current period to determine the predictive value for the subsequent period by using weighted mean. To date, the exponential smoothing method has been widely applied to forecast demand under the bullwhip effect [16] and to plan inventory control strategies [17]. Moreover, the methodology for exponential smoothing has been developed in recent years into one that incorporates the effect of influential factors on the accuracy of demand forecasts [7, 18, 19]. Wang [19] used a model selection method where crucial influential factors were included in the selected model, and nonsignificant factors were removed to avoid over-fitting the model.

Time-series model was first developed in the nineteenth century, and past studies related to such model were then systematically compiled by Box and Jenkins [ 20 ] into a book. A time-series autoregressive integrated moving average (ARIMA) model integrates an autoregressive process and moving average process after obtaining a finite difference from time-series data. The ARIMA model is used to estimate the correlations parameter between the time points of observed values, and the estimated parameter

Table 1. CHARACTERISTICS OF F ORECASTING M ETHODS . (○: YES ; △: YES FOLLOWING

MODIFICATION BY OTHER STUDIES

values can then be used for forecasting. Subsequently, Box and Tiao [21] added other time-series influential factor to the ARIMA model. Pankratz [22] called this model the dynamic regression model. Linear regression models are a type of linear model that are most frequently mentioned in statistical analyses. Linear models assume that research variables and influential factors are linearly related, and thus can be used to explore the effect of influential factors on research variables. Furthermore, linear models assume that observation values are mutually independent; thus, this model is applicable for analyzing data containing mutually independent observation values. If linear models are used to analyze time-correlated data, i.e., the observation values being correlated over time, then unbiased but invalid model coefficient estimators can be obtained. Consequently, the standard errors of the model coefficient estimators would be incorrect, and problems regarding statistical testing within the models arise, such as whether the model coefficients are significantly greater than 0, whether the models exhibit explanatory power, and whether the predictive intervals are reliable in forecast analysis [23, 24]. Linear mixed-effect models can be considered as an extension of linear models. The linear mixed-effect models add random effects to linear models with fixed effects. Hence, a model that has both fixed and random effects is called a linear mixed-effect model. Linear mixed-effect models are typically used to describe the relationship between research variables and categorical factors with correlated observation values. A characteristic of the mixed-effect models is



41

that observation values at the same categorical level have identical random effect values for dependent variables; observation values at different levels have distinct values of random effect. This characteristic explains the correlation between observation values at an identical level. Therefore, linear mixed-effect models differ considerably from linear models. The mixed-effect model can be applied to data where observation values are correlated (e.g., longitudinal data, repeated measures data, and multilevel data). However, linear models can be applied only to data where the observation values are mutually independent. In industrial operations, the pattern of data observations is often time-correlated. For example, when forecasting monthly product demand or monthly inventory levels, the observation values are correlated over time. Under such circumstances, the linear mixed-effect model is more accurate than linear models for identifying statistically significant factors. In the past 2 years, the linear mixed-effect model has been broadly applied in various fields, such as the timber industry [25], medicine [26, 27], and ecology [28], to identify crucial influential factors. In addition, numerous studies have established models for forecasting [29, 30]. However, in industrial engineering and management science [24, 31, 32, 33], no study has used the linear mixed-effect model to make predictions by using time-correlated data or to identify key influential factors. Therefore, in this study, a linear mixed-effect model was applied to business operations to analyze the importance of influential factors, and to forecast product demand; in addition, the performance of the linear mixed-effect model was compared with that of other methods, which are the research contributions of this study. III. LINEAR MIXED-EFFECT MODEL According to parameter attributes, two types of effect exist in a linear mixed-effect model: fixed and random effects [34, 35]. In a linear model, the parameters are all fixed values and therefore its corresponding covariates are referred to as fixed-effect parameters. The fixed effect describes the true value of the coefficient for an entire population, or the true value of the coefficient for a factor that can be repeatedly tested under identical conditions. If a factor in a model exhibits a random effect, then the factor is sampled from an entire population. The random effect is a coefficient of the factor; moreover, the coefficient is a random variable and not a fixed value. The following section introduces the linear mixed-effect model developed by Laird and Ware [36] and the estimation of model parameters, and describes how the research variables are forecasted. A. Linear Mixed-Effect Model In contrast to a multilevel model, a single-level linear mixed-effect model [36] was employed in this study.

The multilevel model differs from the single-level model in terms of the covariance matrix of the observation values. The single-level model involves only one level, whereas the multilevel model involves at least two levels. The covariance matrix of the multilevel model is more complex than that of the single-level model. In practice, whether using a single-level or multilevel model is more appropriate depends on the data structure of the observation values. Although the covariance matrices of the two models differ, the observation values of the various groups at a fixed level are independent of each other, and the within-group observation values are intercorrelated. In the multilevel model, a group at one hierarchy level becomes the next level of the hierarchy.

The single-level linear mixed-effect model developed by Laird and Ware [36] is expressed as follows:

where ib is a matrix that is independent of iε

(index i denotes the ith group at a single level), iy

contains in observation values for the ith group, M

denotes the number of groups, β denotes a p -

dimensional vector for the fixed effect, ib denotes a q -dimensional vector for the random effect, iX

denotes an in p covariance matrix for the fixed

effect, iZ is an in q covariance matrix for the

random effect, and iε denotes an in -dimensional

within-group random error term. The variable iε obeys a multivariate normal distribution with an

expected value of 0 and a covariance matrix of iΛ ,

and ib obeys a multivariate normal distribution with an expected value of 0 and a covariance matrix of Ψ .

The model assumes that iε and jε are mutually

independent ( i j ); in addition, iε and ib are mutually independent. Therefore, considering Models (1) and (2), the covariance matrix of the within-group

observation values iy is expressed as follows: (1)

where the nondiagonal elements of iV are not required to be 0. Therefore, according to (3), Models (1) and (2) allow the existence of the correlation between observation values within a group. This is a major difference that the two models have with the linear model. B. Estimation of the Model Parameters This section introduces estimation methods that adopt the linear mixed-effect model: the maximum



42

likelihood (ML) and restricted ML (REML) estimation methods. Regarding the ML method, the estimates of ML estimators are those that reach the maximum value of ML functions. By comparison, the REML method is aimed at identifying the estimators that exhibit unbiased characteristics. Therefore, estimators obtained using the REML method are unbiased, whereas those derived using the ML method could feature either biased or unbiased property. Therefore, most researchers prefer the REML method [34, 35]. We introduce the estimation procedures for both of these estimation methods, although only the REML method was used in this study. First, the model β coefficient and covariance matrix of observation values iV are estimated as follows. In Models (1) and (2), the expected values of ib and iε are assumed to be 0; thus, the expected value of yi is Xiβ (i.e., ( )i iE y X β ). Because the covariance matrix of yi is Vi (i.e., ( )i iVar y V ) and because ib and iε obey an independent multivariate normal distribution, the marginal distribution of yi is a multivariate normal distribution expressed as follows:

( , )i i iNy X β V� The ML function is expressed as follows:

12

2

1

1

( , )= (2 ) det

1 exp ( ) ( )2

inM

ii

Ti i i i i

L

β θ V

y X β V y X β

where θ denotes the set of 1,..., MV V . To facilitate differentiation, the natural logarithm of the ML function is used instead of the ML function to evaluate the ML and REML estimators, and define l( , ) ln ( , )Lβ θ β θ . ML estimation method The ML estimates of β and θ

are the values that maximize ( , )l β θ and thus are also the values that maximize L(β,θ). Calculating the

maximum value of ( , )l β θ is challenging. Typically,

let ^

θ θ , and evaluate the value of β such that it

maximizes ^ ( , )l θ θ β θ

. Subsequently, let ^

β β , and calculate the value of θ such that it maximizes the

value of ^ ( , )l β β β θ

. This process is iterated until the

change in ^β and

^θ is within a tolerance error (i.e.,

the ^β and

^θ values converge).

Specifically, we first let θ be ^θ (equivalent to

letting iV be ^

iV , 1,...,i M ). Under these

conditions, iy obeys ^

( , )iiN X β V . An analytical solution for β can be obtained by using the

generalized least squares method.

Accordingly, ^

^( , )l θ θ β θ is the maximum value. Next,

fix β in ( , )l β θ as ^β , denoted by ^ ( , )l ββ β θ , to obtain

a θ that maximizes the value of ^ ( , )l ββ β θ , where

where 1,..., MV V are functions of θ . Typically, ^ ( , )l ββ β θ is not a linear function for θ .

Consequently, no analytical solution for θ exists, and an algorithm must therefore be used to obtain a numerical solution for θ . Commonly used algorithms include the expectation-maximization (EM) algorithm, Newton’s method, and Fisher’s scoring algorithm. Previous studies have described these algorithms in detail [36, 37, 38], including a comparison of their strengths and weaknesses [35]. An algorithm can be used to obtain a numerical

solution for θ (i.e., ^θ ), the result of which can be

converted to ^

iV . Subsequently, the calculation is performed iteratively by using Equations (4) and (5)

until the values of ^β and

^θ converge.

REML estimation method The REML method is another approach for estimating θ . The REML estimate of θ is obtained by applying an iterative method to a restricted natural-logarithm ML function.

Regarding the difference between the restricted natural-logarithm ML function (6) and Equation (5), Equation (6) accounts for the loss in degrees of freedom. Therefore, the estimator of θ obtained using the REML is an unbiased estimator. The REML method involves applying Equation (4) to obtain the estimator of β . For the REML, Equations (4) and (6)

are iteratively used until the values of ^β and

^θ

converge. Equation (4) is used in both the ML and REML estimation methods to estimate β . However, the functions employed to estimate θ (i.e., the ML and REML methods use Functions (4) and (6) to estimate θ , respectively) differ between these methods, and



43

they thus yield different values for ^θ . In addition,

because ^

iV is a function of ^θ , different values are

obtained for ^

iV ; consequently, different ^β values are

obtained through using these two methods. Estimating random effect parameters Given ib , the following equation can be derived from (1):

| ( , )di i i i i iN y b X β Z b Λ

where " d " represents "distribution equals" and iΛ is given by (2). Therefore, the generalized least squares method can be applied to estimate ib , which is equal to 1 1 1( ) ( )T T

i i i i i i ii i

Z Λ Z Z Λ y X β . In the

equation, iΛ (a function of θ ) and β are true values. Therefore, by substituting the ML or REML estimates

(i.e., ^β or

^

iΛ ), we can obtain the estimator of ib as follows:

1 1^ ^ ^ ^1( ) ( )T T

i ii i i i i ii i

b Z Λ Z Z Λ y X β .

C. Forecasting Research Variables After the explanatory variables new

iX and newiZ have

been obtained, the estimates of β and ib (i.e., ^β and

^

ib ) described in the previous section can be used to forecast the research variable iy . The predictive value is as follows:

IV. A CASE STUDY This study adopted a single-level linear mixed-effect model to forecast product demand. In the case study, the sample was a leading professional industrial LCD/OLED display manufacturer. This manufacturer produces products that are critical components of various devices used in daily life and are applied in various industries. Moreover, the company has an international customer base. Table 2 shows the number of orders, total product demand, average product demand per order, and quantity of finished goods from 2009 to 2013. Before 2013, the manufacturer produced more than 5,000 product types, and the average quantity of products required in an order was approximately 400. Thus, the manufacturer is considered to be a suitable example of a business that produces a diverse combination of high-mix products. A characteristic of high-mix low-volume manufacturers is that they typically commence production only after receiving a customer order. This

production pattern is typical of the MTO production pattern, which is mainly adopted to serve customers in niche markets. In recent years, the manufacturer’s profits have decreased despite an increasing revenue and market share. Therefore, the manufacturer aimed at changing its production strategy by adopting the MTS production strategy for some product types in order to increase its batch production capacity, reduce its production costs, and improve its production efficiency. In addition, the manufacturer believed that adopting the MTS production strategy would enhance

Table 2. NUMBER OF ORDERS AND PRODUCT DEMAND

customer satisfaction by ensuring the rapid delivery of customer orders, thereby providing a competitive advantage. Thus, being able to accurately forecast product demand was crucial. Following evaluation, to test the implementation of the MTS production strategy, this study selected the top 20 standard finished products that were most frequently ordered between 2011 and 2013 by customers of the sample manufacturer. As shown in Figure 1, these 20 standard products accounted for 20% of the manufacturer turnover for standard products in 2013, with 86 orders placed in the same year. After implementing the MTS production strategy, the manufacturer planned to run production of each product type once per month per year. Accordingly, the production frequency, cost of handling orders, and frequency of changing production lines was reduced. Thus, its long-term production capacity plans can be implemented to maximize the benefits of producing a high volume of products with fewer runs. A. Data Structure The data structure comprised 20 types of standard finished products. The monthly product demand data were collected from January 2007 to December 2013 for each product type (see S1 Table). The historical data before 2012 were used to estimate model parameters, and the model was used to forecast the product demand for 2013 (January–December). Not all 20 products were manufactured from 2007. The historical data used to estimate model parameters comprised 1295 observation values (64 observation values on average for each product type). The product lifecycle varied by year, and the product demand varied by month. Therefore, year and month were crucial predictors. For each type of product, the monthly product demands in each month were related.



44

In this study, the explanatory variables (year and month) were added to the linear mixed-effect model to analyze the monthly product demand data. Regarding product sales, the product demand varied by product type. Accordingly, product type was regarded as a crucial categorical variable because of its influence in forecasting the product demand. In this study, according to the characteristics of the mixed-effect model, we used product type as a random-effect term and included the demand for each product type in a universal model to

Figure 1. Maximum total number of orders (2011–2013). The plot shows that the accumulated percentage of the maximum total number of orders from 2011 to 2013 is less than the turnover of standard products in 2013. The first 20 products accounted for approximately 20% of the turnover for of standard products. The numbers in green denote the number of orders for standard products in 2013 corresponding to the horizontal axis. forecast the demand for type separately. Subsequently, we compared other commonly used forecasting methods. Unlike the mixed-effect model, other methods did not have a universal model to account for 20 unique product types. Therefore, for the other forecasting methods, the data are required to be divided into multiple data sets according to product type, and the partitioned data are then applied to the forecasting methods depending on the product type for analysis and forecasting. This approach substantially reduces the sample size, reducing the accuracy of the forecast. B. Model Development Product demand differed by product type, and thus we assumed the demand for each type of product to be mutually independent. In Model (1), which is the single-level model, random effect was set to be product type, thus yielding various random-effect coefficient for each product type. The model is expressed as follows:

where iy is a vector that denotes the monthly product demand (the vector length is equal to the data quantity for product i); β0, β1, β2, and β3 denote the intercept, year, year-squared, and month for the fixed-effect term; and 0ib and 1ib denote the intercept and year-squared for the random-effect term. In Model (8), year was considered as a continuous variable with 2007 used as the baseline. Month was a categorical variable; therefore, the month term in Model (8) was a dummy variable. The dummy variable for month had 11 indicator variables with a value of 0 or 1, and the total product demand in January was used as the baseline. Expressing Equation (1) as Model (8), the fixed-effect explanatory variable iX is a matrix comprising a column of 1’s vector for the intercept, year, year-squared, and month covariates. Thus, the expression 0 1 2 3=[ ]T T β β is a 14 × 1 vector, where 3β is the coefficient of the dummy variable for the month covariate and has 11 elements. To account for the various product types, we chose the intercept and year-squared covariate as the random-effect explanatory variable, where the intercept was used to account for the average difference of demands between product types, and the year-squared covariate was used to consider the difference between product demands decreased or increased over time. The explanatory variable iZ in the random-effect explanatory variable comprised the intercept and year-squared covariate, of which the coefficients are a 2 × 1 vector expressed as 0 1=[ ]T

i i ib bb . In Model (8), the year-squared covariate in the random-effect explanatory variable was also a part of the fixed-effect explanatory variable, and was used to account for the fact that the expectation of ib was probably unequal to 0; thus, the assumption that ib in (2) was equal to 0 was reasonable. The year-squared covariate was included to prevent the annual growth trend from being linear, which enabled the model to more accurately reflect the current situation. The year-squared covariate is crucial to practical operations. The year and year-squared covariates added into the fixed-effect explanatory variable facilitated establishing a grand model for the 20 product types. The year and year-squared covariates for the fixed effect indicated the average growth trend for the 20 product types, whereas the random effect reflected the specific annual growth trends for each product type. To forecast the monthly product demand for 2013, 2013 was used as the value for the year and year-squared covariates. Both covariates and the target month were input into the explanatory variable to

form newiX and new

iZ . Subsequently, ^β and

^

ib in (7)

were used to obtain the forecasted value ^

iy .



45

C. Other Forecasting Methods Comparing forecasting methods is crucial in methodological studies [39 , 40 , 41 , 42 , 43]. The model proposed in this study was compared with commonly used statistical forecasting methods, beginning with the following linear model:

Table 3. Linear Mixed-Effect Model Versus the Linear Model.

“ - ”: p < .1; “ * ”: p < .05; “ ** ”: p < .01; “ *** ”: p < .001.

where 0 , 1 , 2 , and 3 are regression coefficients and 3 denotes the coefficient of the dummy variable for the month covariate, and j is the error term. Model (9) (i.e., the linear model) includes only the fixed-effect term in Model (8) (i.e., the mixed-effect model); therefore, Model (9) was compared with Model (8) to examine the differences when the random-effect term is present or absent in the model. A total of 1295 observations of monthly product demand ( jY , 1,...,1295j ) were used to estimate the coefficients in Model (9) and the significance of the coefficients with P values. In the Results section, Models (8) and (9) are compared regarding forecast accuracy and the P values. Next, the model proposed in this study was compared with the exponential smoothing method, in which the product demand observation values tY ’s and its predictive values tF ’s were used to obtain the predictive values for the subsequent period by calculating a weighted mean. The forecast formula is as follows:

1 (1 )t t tF Y F

where is the weighted coefficient. To accurately forecast the monthly product demand in this case, we adjusted the exponential smoothing method to account for two influential factors (i.e., month and product

type). The data were divided into 20 data sets according to each product type, and each data set was divided into 12 subsets (one for each month). For each product type, no more than six observations from each month in the historical data were used. The pre-2012 monthly product demand data were used to forecast the product demand for the corresponding months in

2013. The weighed coefficient was 12( 1)N

,

where N is the number of observations for a month (N ≤ 6). Finally, the model proposed in this study was compared with a seasonal time-series model; specifically, the autoregressive moving average model (ARMA(2,2)12), which was considered to be a suitable model because the data were not nonstationary time-series data. The mathematical model for ARMA( , )sp q is expressed as follows:

1 1(1 ) (1 )

p qs i s i

i t i ti i

B Y B

where i is the ith order autoregressive process coefficient, B is a backward shift operator, i is the ith order moving-average process coefficient, t is a normally distributed confounding term, and s is a seasonal parameter. Longitudinal data were collected for each of the 20 product types. A time-series model was established for each of the 20 product types. In this case, the month was regarded as a crucial influential factor for forecasting and thus the seasonal parameter s was set to 12, which indicates the existence of correlations in the data for every 12 month. The samples were categorized by product type, yielding an average of 64 samples for each type of product. The parameters p and q were determined based on the characteristics of an autocorrelation function, a partial autocorrelation function, and an extended autocorrelation function (p = 2 and q = 2). Finally, the ARMA(2,2) 12 model was used to forecast the product demand for each product type.

Table 4. Error Indicators for the Four Forecasting Methods.



46

D. Results In this study, mean of absolute error (MAE), mean of absolute percent error (MAPE), and root-mean-square error (RMSE) were used as error indicators. The definitions for these error indicators are provided as follows:

1

1M A E | |

n

t tt

n F Y

1

1M A P E 1 0 0 | |

nt t

t t

F Yn

Y

0 . 51 2

1R M S E ( )

n

t tt

n F Y

where n denotes the number of months to be forecasted (n = 12 in this case), Yt represents the true product demand for month t of 2013, and Ft is the forecasted product demand for month t. The fixed-effect term in the linear model was compared with that in the linear mixed-effect model. As shown in Table 3, the absolute values of the coefficients for the explanatory variables in the linear mixed-effect model containing the random-effect term are greater (i.e., further from 0) than all of those in the linear model except for April. In addition, the standard errors and P values for all of the explanatory variables in the linear mixed-effect model are smaller than those in the linear model. Regarding the linear fixed-effect model, compared with January in a given year, the product demand was significantly greater in May and November (P value < 0.1), in July (P value < 0.05), and in March, April, and September (P value < 0.01). Compared with the linear fixed-effect model, the linear model yielded less significant results. The linear model is suitable for data containing mutually independent observation values. In this case, the observation values for product demand were correlated over time, thereby violating the assumption of the linear model. Therefore, the standard errors and P values for the linear model (Table 3) are not valid estimates, whereas those for the linear mixed-effect model are more reliable. Table 4 shows the error indicators for the four forecasting methods. Because this case involved three error indicators for each of the 20 product types, Table 4 presents the mean and standard deviation of the three error indicators. As shown in Table 4, the means and standard deviations of MAE, MAPE, and RMSE for the linear mixed-effect model are lower than those for the linear, ARMA, and exponential smoothing models, indicating that, in this case, the linear mixed-effect model is superior to the other three models. Regarding the model comparison (Table 5), the predictive values obtained through using the linear model to process the correlated data are unbiased [23]. However, the linear mixed-effect model (8) contains the random-effect term, whereas the linear model (9) does not. Therefore, in Model (8), the intercept and year-squared terms differ according to the product type, and thus the corresponding intercept values and coefficients differ based on the product type. In Model (9), the covariate of product type is not included in the explanatory variables, which generates

identical predictive values for various product types in the same years and months. Thus, this model cannot predict the product demand for the individual product types, rendering its forecasting effectiveness inferior to that of Model (8). Regarding the exponential smoothing method, we considered product type and month as crucial influential factors, which were used as the basis for dividing the data into 240 data sets. For each product type, the pre-2012 monthly data were used to forecast the monthly product demand for 2013. In this manner, the exponential smoothing method was applied 12 times for each of the 20 product types. In addition, less than six observations from the historical data were used in the exponential smoothing method (for a given month, there were at most 6 sets of data from 2007 to 2012); consequently, the risk of inferential error was high because only a few observations were involved in the prediction. Regarding the seasonal time-series model ARMA(2,2)12, we considered product type as a crucial influential factor and divided the data into 20 data sets according to product type. For each product type, 64 observations were used on average. The ARMA(2,2)12 model was used to forecast the product demand for each product type by considering the correlation between the data for every 12 month. For both the exponential smoothing method and the ARMA(2,2)12 model, the data were divided into subsets according to the product type and then used to estimate the monthly effect of each product type. Accordingly, although such procedure could consider the various monthly effects for various product types and the interaction between product type and month, it reduces the number of data observations involved in the prediction. In the linear mixed-effect model, 1295 data observations were used to estimate the random effect for each product type. The number of data observations used in the linear mixed-effect model was considerably more than that used in the exponential smoothing and time-series models, which could explain

Table 5. Comparison of the Four Models.

※1 This effect is nonsignificant

why the linear mixed-effect model produced lower forecast errors. In addition, in Model (8), the random effect of the interaction term for month and year-



47

squared term was considered and the likelihood ratio test was employed to examine whether this term is significant to this model. The results showed that only the random effects of the intercept and year-squared terms were significant, and the random effect of the month term did not significantly enhances its explanatory power for the data. Therefore, the random effect of the interaction term was not included in Model (8). V. DISCUSSION In summary, when applying the linear mixed-effect model, all of the historical data were used in one model to predict the monthly product demand for each product type, and to avoid problems resulting from dividing the data into smaller data sets. In this case study, using the linear mixed-effect model enables manufacturers who adopt the MTS production strategy to predict the amount of inventory they should stock. Furthermore, the model is more effective in forecasting product demand than is the time-series, exponential smoothing, and linear models. Similar to the linear model, the linear mixed-effect model is typically used to examine the relationship between explanatory and research variables. Unlike the linear model, which assumes the observation values to be mutually independent, the linear mixed-effect model is suitable for examining correlated data. Because the data pertaining to business operations are generally correlated over time, the linear model is limited in applicability. By contrast, the linear mixed-effect model was initially developed to handle correlated data. Other methods such as the time-series and exponential smoothing methods formulate the correlation between observation values as parameters, and then estimate the parameters by data and forecast the observations by the estimates. When the time-series and exponential smoothing models were first developed, these methods were not aimed at analyzing the relationship between explanatory and dependent variables. Wang [19] proposed an exponential smoothing method that included explanatory variables and can be used to explore the association of research variable. Because this method is a relatively new development, most of statistical software packages have not yet incorporated related functions, and thus this method has not been widely used. By contrast, the linear mixed-effect model was developed more than 30 years ago, and related functions have been included in various statistical software packages. Using linear mixed-effect, time-series, and linear models to forecast product demand can yield negative predictive values. This phenomenon occurs when the linear mixed-effect model is used because iε in (2) is assumed to be normally distributed and the link function is an identity function. Negative values are

usually obtained from historical data where product demand is zero or very low. To prevent this, predictive value was truncated at 0 (i.e.,

^max( , 0)ttF Y , where

^

tY denotes a predictive value derived from any method, and Ft denotes an actual predictive value obtained from any prediction

method). In other words, if ^

0tY , then ^

ttF Y ; if ^

0tY , then 0tF . Some link functions in generalized linear mixed-effect model can deal with the case where dependent variable is restricted to

^0tY [44]. However, the prediction intervals for the

random-effects in linear mixed-effect model are well developed [45, 46, 47, 48, 49]. It is useful to apply the prediction intervals in business operations for knowing whether the random-effect exists. Implementing an MTS production strategy can enhance the competitive advantages of a manufacturer, enabling the manufacturer to rapidly satisfy product demand, thereby reducing internal and external transaction costs for handling orders. Employing this strategy also enables high batch centralized production and thus can reduce production costs and assist manufacturers in negotiating with material suppliers about the cost of materials. Because this approach enables short delivery times, customer satisfaction can be improved, thus attracting potential customers who need products immediately. Consequently, market share can be increased. MTS production also enhances the usage rate of production equipment. Companies that adopt an MTS strategy require an accurate forecasting method to realize these advantages. This study proposed an accurate forecasting method for determining the stock levels a company should determine for adopting the MTS production strategy, a topic that has seldom been discussed in studies on MTS production. Using an MTS production strategy involves the potential risk of increasing inventory costs. Therefore, future studies should adequately apply the strengths of the linear mixed-effect model (e.g., accurately forecasting demand for multiple product types in one go) when forecasting. Future studies should consider investigating whether the forecasting intervals of the linear mixed-effect model can be coupled with various inventory strategies to assist manufacturers with adopting the MTS production strategy in order to develop an optimal business operation model in terms of optimal inventory time points and minimal inventory costs. In addition, to remain competitive, companies should enhance their organizational capability for elevating the threshold that enables competitors to develop similar operating models. Future studies are also recommended to explore the benefits that the MTS production strategy involving a linear mixed-effect model brings to the various



48

departments of an enterprise and the effects of such strategy on customer satisfaction and loyalty. REFERENCES

[1] P. Danese and M. Kalchschmidt, "The role of the forecasting

process in improving forecast accuracy and operational performance," International Journal of Production Economics, vol. 131, no. 1, pp. 204-214, 2011.

[2] I. J. Adan and J. Van der Wal, "Combining make to order and make to stock," Operations-Research-Spektrum, vol. 20, no. 2, pp. 73-81, 1998.

[3] T. Williams, "Special products and uncertainty in production/inventory systems," European Journal of Operational Research, vol. 15, no. 1, pp. 46-54, 1984.

[4] W. M. Guillot, "Strategic leadership: Defining the challenge," Air and Space Power Journal, vol. 4, 2003.

[5] R. R. Magee, "Strategic leadership primer," DTIC Document, Tech. Rep., 1998.

[6] S. A. Shambach, "Strategic leadership primer," DTIC Document, Tech. Rep., 2004.

[7] R. Blackburn, K. Lurz, B. Priese, R. Gob, and I.-L. Darkow, "A predictive analytics approach for demand forecasting in the process industry," International Transactions in Operational Research, 2014.

[8] C. A. Soman, D. P. Van Donk, and G. Gaalman, "Combined make-to-order and make-to-stock in a food production system," International Journal of Production Economics, vol. 90, no. 2, pp. 223–235, 2004.

[9] W. Popp, "Simple and combined inventory policies, production to stock or to order?" Management Science, vol. 11, no. 9, pp. 868–873, 1965.

[10] L. C. Hendry and B. Kingsman, "Production planning systems and their applicability to make-to-order companies," European Journal of Operational Research, vol. 40, no. 1, pp. 1–15, 1989.

[11] S. Rajagopalan, "Make to order or make to stock: model and application," Management Science, vol. 48, no. 2, pp. 241–256, 2002.

[12] M.-K. Chen and A. Cheng, "The study of supply chain inventory strategy under bto production environment," Journal of the Chinese Institute of Industrial Engineers, vol. 20, no. 4, pp. 398–410, 2003.

[13] L. Breiman., "Statistical modeling: The two cultures (with comments and a rejoinder by the author)," Statistical Science, vol. 16, no. 3, pp. 199–231, 2001.

[14] C. C. Holt, "Forecasting seasonals and trends by exponentially weighted moving averages," International Journal of Forecasting, vol. 20, no. 1, pp. 5–10, 2004.

[15] J. F. Muth, "Optimal properties of exponentially weighted forecasts," Journal of the american statistical association, vol. 55, no. 290, pp. 299–306, 1960.

[16] F. Chen, J. K. Ryan, and D. Simchi-Levi, "The impact of exponential smoothing forecasts on the bullwhip effect," Naval Research Logistics, vol. 47, no. 4, pp. 269–286, 2000.

[17] S. C. Graves, "A single-item inventory model for a nonstationary demand process," Manufacturing & Service Operations Management, vol. 1, no. 1, pp. 50–61, 1999.

[18] R. Gob, K. Lurz, and A. Pievatolo, "Electrical load forecasting by exponential smoothing with covariates," Applied Stochastic Models in Business and Industry, vol. 29, no. 6, pp. 629–645, 2013.

[19] S. Wang, "Exponential smoothing for forecasting and Bayesian validation of computer models," Ph.D. dissertation, Georgia Institute of Technology, 2006.

[20] G. E. Box, "Time series analysis: Forecasting and control,," Time Series and Digital Processing, 1976.

[21] G. E. Box and G. C. Tiao, "Intervention analysis with applications to economic and environmental problems," Journal of the American Statistical association, vol. 70, no. 349, pp. 70–79, 1975.

[22] A. Pankratz, "Forecasting with dynamic regression models," Wiley Series in Probability and Mathematical Statistics.

Applied Probability and Statistics, New York: Wiley, 1991, vol. 1, 1991.

[23] L. C. Hamilton and D. Press, Regression with graphics: A second course in applied statistics. Duxbury Press Belmont, CA, 1992, vol. 1, no. 1.

[24] J. G. De Gooijer and R. J. Hyndman, "25 years of time series forecasting," International journal of forecasting, vol. 22, no. 3, pp. 443–473, 2006.

[25] P. Hiesl and J. G. Benjamin, "Estimating processing times of harvesters in thinning operations in maine," Forest Products Journal, 2014.

[26] M. L. Bourbonnais, T. A. Nelson, M. R. Cattet, C. T. Darimont, G. B. Stenhouse, and D. M. Janz, "Environmental factors and habitat use influence body condition of individuals in a species at risk, the grizzly bear," Conservation Physiology, vol. 2, no. 1, p. cou043, 2014.

[27] Y.-C. Tsai, Y.-W. Chiu, J.-C. Tsai, H.-T. Kuo, S.-C. Lee, C.-C. Hung, M.-Y. Lin, S.-J. Hwang, M.-C. Kuo, and H.-C. Chen, "Association of angiopoietin-2 with renal outcome in chronic kidney disease," PloS one, vol. 9, no. 10, p. e108862, 2014.

[28] T. G. Holmes, W. M. Tonn, C. A. Paszkowski, and G. J. Scrimgeour, "Effects of winter surface aeration on pelagic zooplankton communities in a small boreal foothills lake of alberta, canada," Journal of Freshwater Ecology, no. ahead-of-print, pp. 1–14, 2014.

[29] N. Mohsin, G. Mourad, M. Faure, I. Szawarc, and J. Bringer, "Metabolic syndrome performs better than the individual factors in predicting renal graft outcome," in Transplantation proceedings, vol. 45, no. 10. Elsevier, 2013, pp. 3517–3519.

[30] A. Schwalm, Y.-S. Feng, J. Moock, and T. Kohlmann, "Differences in eq-5d-3l health state valuations among patients with musculoskeletal diseases, health care professionals and healthy volunteers," The European Journal of Health Economics, pp. 1–13, 2014.

[31] Y. Acar and E. S. Gardner, "Forecasting method selection in a global supply chain," International Journal of Forecasting, vol. 28, no. 4, pp. 842–848, 2012.

[32] P. G. Allen and B. J. Morzuch, "Twenty-five years of progress, problems, and conflicting evidence in econometric forecasting. what about the next 25 years?" International Journal of Forecasting, vol. 22, no. 3, pp. 475–492, 2006.

[33] R. Fildes, "The forecasting journals and their contribution to forecasting research: Citation analysis and expert opinion," International Journal of forecasting, vol. 22, no. 3, pp. 415–432, 2006.

[34] J. C. Pinheiro and D. M. Bates, Mixed-effects models in S and S-PLUS. Springer Science & Business Media, 2000.

[35] B. T. West, K. B. Welch, and A. T. Galecki, Linear mixed models: a practical guide using statistical software. CRC Press, 2014.

[36] N. M. Laird and J. H. Ware, "Random-effects models for longitudinal data," Biometrics, pp. 963–974, 1982.

[37] N. Laird, N. Lange, and D. Stram, "Maximum likelihood computations with repeated measures: application of the EM algorithm," Journal of the American Statistical Association, vol. 82, no. 397, pp. 97–105, 1987.

[38] M. J. Lindstrom and D. M. Bates, "Newton Raphson and EM algorithms for linear mixed-effects models for repeated-measures data," Journal of the American Statistical Association, vol. 83, no. 404, pp. 1014–1022, 1988.

[39] P.-F. Pai, K.-P. Lin, and J.-S. Wang, "Stock price forecasting in Taiwan using ellipsoidal fuzzy system," Journal of the Chinese Institute of Industrial Engineers, vol. 21, no. 2, pp. 146–155, 2004.

[40] W.-Y. Hwang and J.-S. Lee, "A new forecasting scheme for evaluating long-term prediction performances in supply chain management," International Transactions in Operational Research, vol. 21, no. 6, pp. 1045–1060, 2014.

[41] Y. Cao, G. Wan, and F. Wang, "Predicting financial distress of Chinese listed companies using rough set theory and support vector machine," Asia-Pacific Journal of Operational Research, vol. 28, no. 01, pp. 95–109, 2011.

[42] F. B. e Silva, E. Koomen, V. Diogo, and C. Lavalle, "Estimating demand for industrial and commercial land use



49

given economic forecasts," PloS one, vol. 9, no. 3, p. e91991, 2014.

[43] X. Zhang, T. Zhang, A. A. Young, and X. Li, "Applications and comparisons of four time series models in epidemiological surveillance data," PloS one, vol. 9, no. 2, p. e88075, 2014.

[44] J. Jiang, Linear and generalized linear mixed models and their applications. Springer Science & Business Media, 2007.

[45] D. R. Cox, "Prediction intervals and empirical Bayes confidence intervals,” Perspectives in Probability and Statistics, pp. 47–55, 1975.

[46] C. Morris, “Parametric empirical Bayes inference: theory and applications,” Journal of the American Statistical Association, vol. 78, no. 381, pp. 47–55, 1983.

[47] R. Basu, J. K. Ghosh, and R. Mukerjee, “Empirical Bayes prediction intervals in a normal regression model: higher order asymptotics,” Statistics & probability letters, vol. 63, no. 2, pp. 197–203, 2003.

[48] S. Chatterjee, P. Lahiri, and H. Li, “On small area prediction interval problems,” The Annals of Statistics, vol. 36, pp. 1221–1245, 2008.

[49] M. Yoshimori, and P. Lahiri, “A second-order efficient empirical Bayes confidence interval,” The Annals of Statistics, vol. 42, no. 4, pp. 1–29, 2014.


50

MICRO ENCAPSULATION OF PIPER BETLEL. AND BOESENBERGIA PANDURATAESSENTIAL OILSUSING INCLUSION

COMPLEXATION AND THEIR ANTIMICROBIAL ACTIVITY

1BUSSABA POOSRI, 2RACHA TEPSORN, 3CHUENJIT CHANCHAROONPONG, 4SURIYAN SUPAPVANICH

1, 2Division of Food science and Technology TU,

3Faculty of Natural Resource and Agro-IndustryKU, Agricultural Education Department KMITL E-mail: [email protected], [email protected],[email protected],[email protected]

Abstract - Boesenbergia pandurata essential oil(BEO)and Piper betleL. essential oil(PEO)were encapsulated in -CD by co-precipitation method. The results indicated that the initial essential oil content and type of EO had an effect on the physico-chemical and antimicrobial properties. The PEO microparticles were more effective than the BEO microparticles. The best ratios of the BEO and PEO/-CD inclusion complex was a 50:50 with 98.88 (%EE), 99.51 (%EC), 63.20 (%yield) anda 25:75 with72.80(%EC), 35.36 (%EC), 64.93 (%yield) respectively.The morphology analysis and the encapsulation analysis of themicroparticles were confirmed by scanning electron microscopy. All images revealed irregular shape and strong tendency of agglomeration. The BEO microparticlesand PEO microparticles showed inhibitory and bactericidal effects on six tested bacteria. The PEO/-CDinclusion complex showed the lowest concentration of MIC (0.15%w/v) on B. cereusATCC 11778, S.TyphimuriumATCC 13311and Y.enterocoliticaDMST 17911. While the BEO/-CDinclusion complex displayed the lowest concentration of MIC(0.3%w/v) on E. coli 6795 and S.TyphimuriumATCC 13311. This finding suggested that BEO andPEOinclusion complexes could be applied tonatural food preservative in food industries. Index Terms - Inclusion complexation, Boesenbergia pandurata essential oil, Piper betleL.essential oil, Microorganism -cyclodextrin, Antimicrobial activity I. INTRODUCTION Essential oils(EOs)from plant arevolatile oily liquid of secondary metabolism and showed antimicrobial properties (Angioni, et al., 2004; Rahman and Kang, 2009)and are categorized as GRAS (Generally Recognizedas Safe)by the US Food and DrugAdministration (Manso, et al., 2013). Boesenbergia panduratais an annual plant of Zingiberaceaeand Piper betle L. is an annual plant ofPiperaceae family.They are widely used as natural spice and medicine in Thailand(Puangpen, et al., 2008; Arunratet., al 2009). Cyclodextrin(CDs) are cyclic oligosaccharides derived from starch composed of 6, 7 or 8 (α, and γ-CD) glucopyranoseunits, with a relative hydrophobic interior space andhydrophilic exterior surface. The encapsulation of EOs wasbased on the inclusion hydrophobic molecules in the interior space of polymer with the forming of a water soluble inclusion complex(Wang, et al., 2011; Hill, et al., 2013) Previous studies reported that phenolic structure in essential oils are mainly responsible for their biological activity (Burt, 2004; Bakkali, et al., 2008).However, the problem of essential oils which as volatile compound were insoluble-water agent and sensitive to light, oxygen and heat in environment condition. To overcome this drawback, -cyclodextrins(-CDs)were used to encapsulate the essential oils toprotect them from the environment, to improve aqueous solubility, and to preserve

bioavailability.The aim of this study is to evaluate the physico-chemical and their antimicrobial properties of Boesenbergia pandurataessential oilcomparedtoPiper betleL.essential oilsin -CDby co-precipitation method. II. MEDTERIALS AND METHOD A. Materials -cyclodextrin(fromBrenntag,Germanymolecular mass of 1,135 Da, purity > 95%) and essential oil in this study, which include Piper betle L.(PEO)andBoesenbergia pandurata(BEO)were provided by Thai-China Flavoursand Fragrances Industry Co.,Ltd. (Ayutthaya, Thailand),purity were more than 95 %. B.Preparation of Inclusion complexation with -cyclodextrin(-CD)

Fig.1 Schematic of the principle of inclusion complexation

in -CD The preparation of the EO/-CDcomplex was determined by modification as Bhandari et al.(1998)using the co-precipitation method.To prepare-CDsolution,4.5 g of-CD was dissolved in 50 mLof water: ethanolsolution (2:1), at 55 ºC. PEO and

Micro Encapsulation of Piper Betlel. and Boesenbergia Pandurata Essential Oils using Inclusion Complexation and their Antimicrobial Activity


51

BEO were taken in different quantities and added into-CD solution, in order to obtain a mixture with a ratio (w/v): 50:50; 40:60; 30:70 and 25:75. Samples were stirred by an electromagnetic stirrer for 30 minutes at 55 ºC and were continuouslystirred for4 hours without heating. The mixtures were kept at 4 ºC for 24 hours. The precipitation of the complex mixture was separated through Whatmanfilter paperno.4 and was dried in a hot air oven at 55 ºC for 24 hours. The encapsulated EO/-CDpowder was kept in a desiccator with calcium chlorite overnight,in order to absorb the remaining moisture from the powder. Finally, the powder was preserved in a grown bottle at room temperature. C.Scanning electron microscopy (SEM) The physical structural analysis of microparticles was carried out though scanning electron microscopy(JEON, model no. JSM-7800F, Tokyo, Japan). In orderto record SEM micrographs, samples of the microcapsule were attached to aluminum stubs with a double-sided adhesive tape and coated with gold by a sputter-coater. Then, SEM images were examinedat 4000xmagnification using an acceleration voltage of 1.5 kv. D.Encapsulation efficiency, encapsulation capacity and yield of microparticles The method described by Bhandari et al. (1998) was used with a modification and was carried out through a washing method.To determine the total content of essential oilin encapsulated powder. 0.5 g of powder were suspended in mixed solution with 8 mLdistilled water and 8 mL hexane which was then stirredat 70 ± 2ºC for 20 minutes. -CD was separated from the extract through Whatmanno.4 filter paper and washed tree times with 2 mL of n-hexane. After that, the water was removed out from the extract through a nitrogen gas spitting. The oily residue was then weighted. In order to determine initial essential oil content was absorbed on the surface of a microcapsule. 20 of mL hexane was added ontoa 0.5 g sample and then stirred with an electromagnetic stirrer for 20 minutes without heating. The sample was filled by filter paper to separate -CD from the essential oil and was taken out to remove water in solvent through nitrogen gas. Finally, the oily residue was then weighted. The encapsulation efficiency(%EE) was calculated with the equation (1)(Klinkeson, et al, 2006).

Where E1 is the total content of EO encapsulated in the cavity of -CD plus the oil content absorbed on surface of -CD molecules and E2 is the oil content absorbed on the surface of microcapsule. The encapsulation capacity (%EC) was calculated with the equation (2) (Pudukka, et al, 2000).

Where S0 is the g of initial content of the EO sample used for encapsulation. The total yield value (%yield) was calculated with the followingequation (3): Where y0 is the initial content of Piper betleL. EO or Boesenbergia pandurata EO used for encapsulationplus the initial content of -CD used as a coating and y1is the final oil content plus the final -CD content remainsof microcapsule. E. Antimicrobial activity of microparticles E-1. Bacterial strains The test microorganism were obtained from Department of Medicine Science (DMSC, Thailand), the six strains of bacterial consist of Escherichia coli 6795,SalmonellaTyphimurium ATCC 13311, Yersinia enterocolitica DMST 17911, Pseudomonas aeruginosa ATCC 27853Bacillus cereus ATCC 11778 and Staphylococcus aureus ATCC 25923, which were representative of foodborne pathogens and food-related microorganisms.Allstains were kept in acryotube and TSA slant agar. Tubes were maintained at -18 ºC and 4 ºC in refrigerators respectively. E-2. Determination of the Minimum Inhibitory and Bactericidal concentration (MIC/MBC) The Minimum inhibitory concentration (MIC)of the microcapsule BEO and PEO were determined against different microbialsby macro broth dilution methodand a modification as described by DVG method: (DVG, 2000).To prepare a stock solution, the microcapsule EO was dissolved intosterile distilled water and was serially diluted at concentrations of 0(control), 20, 10, 5, 2.5, 1.25, 0.625, 0.3125 %w/v.2 mL of the stock solution of each concentration was added to test tubes containing 2 mL of TSB× 2to adjust the final concentration range of from 0.15625 % -10 % w/v respectively andfinal volume as 4 mL.After that, the test tubes were homogenized by a vortex mixer for 1 minutes. For the control sample, sterile distilled water (2 mL) was addedinto the test tubewithout antimicrobialmicrocapsule. Then,a 0.1 mL suspension (1×106 CFU/mL)of tested bacteria was added to each tube. Finally, the test tubes were incubated overnight at 37 ºC for 24 hours. The lowest concentration of the microcapsule EO showed complete inhibition of growth of each tested bacteria was defined as MIC values. All test tubes that showed inhibition were assessed through bactericidal capability (Minimum Bactericidal Concentration)by a loop fulstreaked on TSA plates and incubated at 37 ºC for 24 h. If no bacteria’s growth was observed on the plate surfaces following incubation, the treatment concentration was considered as bactericidal. The lowest concentration of antimicrobial indicating bactericidal activity across all replicates was considered as MBC values.



52

F.Statistical analysis Data analysis was performed by using SPSS statistical software version 16 (SPSS, Chicago, IL). All experiments were carried out in triplicateand the data were used to calculate the mean ± standard deviation. Analysis of variance (One way-ANOVA) followed byFisher's least significantdifference (LSD)procedure was used to identify any significant difference between each sample at a 95 % confidence level. III. RESULSTS AND DISCUSSION A.Morphology of BEO/PEOinclusion complexes microparticles Fig.2 shows the SEM micrographs of -CD without loaded essential oil(Control).From this image, it can be observed that unload -CD was non-crystalline and irregular in shape and with a smooth surface due to the absence of the inclusion complexation process between essential oil with -CD.

Fig.2 SEM micrograph of essential oil unload -CD at 4000×

magnification represented as control

Fig.3 shows the SEM micrographs of BEO/-CD complex particles by co-precipitate method at different ratios. The micrographs revealed that different ratio have effect on the particle size and the crystallization form.

Fig.3 SEM micrographs of the encapsulated Boesenbergia

pandurataEO in -CD complex at different ratios; 50:50 (a), 40:60 (b), 30:70 (c) and 25:75 (d) with 4000× magnification

From this figure, it can be noted that all of these particles show irregular shapes and sharp edges with

significant differences in size distribution compared to unload /-CD (Fig.2). In addition, these images also reveal evidence of agglomeration of particle. Similar results was reported by Seo at al. 2010 that theeugenol/-CD complex formed large aggregatesuggested that their irregular form was aconsequenceof the self-assembly of -CD in water.Due to a lack of significant net charge on the inclusion complex particle, this resulted inno repulsive forces to prevent particle agglomeration.(Hill, et al. 2013). Comparing of SEM micrographs at the same magnitude (×4000)revealed that using a BEO/-CD in a 50:50 (%w/v) ratio has larger particle size of microcapsule than the other ratioand 40:60 ratio has particle size and morphology close to this ones, while 30:70 ratio and 27:75 ratio have clearly separate particlesand smaller particles as shown in Fig. 3(c) and Fig. 3(d).This result demonstrated that when the initial content of essential oil used for encapsulation increased, the particle size will increase respectively due to encapsulationcapacity increasing. The different size and agglomeration as a function of initial EO content corresponds to previous studies of the loading of coriander essential oil in -CD microparticles (Dima,et al, 2014). Additionally, similar resultswere reportedby Choi et al, 2010 in the encapsulation of fish oil in -CD complex.

Fig. 4 SEM micrographs of the encapsulated Piper betleLinn. EO in -CD complex at different ratios; 50:50 (a), 40:60 (b),

30:70 (c) and 25:75 (d) with 4000× magnification

Fig.4 showsthe size and morphology of PEO/-CD complex particles at different ratios. Themicrographs revealed irregular shapes with sharp edges of microcapsule and showed different size distribution when compared with its control (Fig.2).Besides, the micrograph showed agglomeration of particle similar to BEO/-CD.From these images, it can be observed that at a 25:75 ratio,BEO/-CD inclusion complexhas larger particle size than the other treatments while 30:70 ratio, 40:60 ratio and 50:50 ratio have smaller particle size respectively. This result indicated that as



53

theinitial essential oilswere used for encapsulation increased, the particle size of the microcapsuledecreased because of the differencesin ratio.This resultcontrasts with BEO/-CDcomplex because the differentratio and the main structure of their active compoundmolecules in essential oil have influence of access into the cavity of -CD. B.Effect of encapsulation efficiency, encapsulation capacity and yield of BEO and PEO/-CDinclusion complexes The encapsulationefficiency(EE), encapsulation capacity (EC)and yield value of BEO/-CDinclusion complexesby co-precipitation method were calculated using Eq. (1), Eq. (2)and Eq. (3) andare listed inTable 1.

Table 1 Encapsulation efficiency, encapsulation capacity and yield values of BEO/-CD inclusion complex microparticles

a-dMean with different subscripts in the same column are significantly different (P <0.05). The data presented tree times

experiment. The results showed that the encapsulation efficiency, the encapsulation capacity and the yield value of microparticleswere all influenced by the initialessential oil content. The EEs of the BEO/-CD inclusion complexes ranged from 93.9605 g/100 g to 98.8817 g/100 g, and the ECs ranged from 53.4934 g/100 g to 99.5131 g/100 gwhich indicated the highlyeffective encapsulation of the essential oil by-CDcarrier.The yield value of the BEO/-CD inclusion complexes ranged from 63.2083 g/100 g to 80.3012 g/100 g.When compared EEvalues ofthe BEO/-CD inclusion complexes at different ratiosfound that a 50:50 ratio showed the highest encapsulationefficiency (98.8817 ± 0.52),and was significantlydifferent(P<0.05)to other ratios.Moreover,a 50:50 ratioalso has thehighest encapsulation capacity (99.5131 ± 0.69) among theratios. The results demonstrated that the encapsulation efficiency and capacity values increased according tothe increasedinitialessential oil content. The effectiveness of the BEO/-CD complex microparticles were due to the suitability of materials and coating. In addition, it possible to resulted from mainly componentof BEO, which consist of 1, 3, 7-octatriene, 3, 7-dimethyl, geranial, camphor and camphene (Rattanakreetakulet al. 2005). They are small molecularand short

chain.Thus, they are easily accessible in the cavity of -CD. When considering the yield of microparticles at different ratiosfound that the lowest loading has highest yield value. This observation suggests that when the essential oil loading increased, the yield value was dropped because the absorption of essential oilsonthe surface of-CD decreased.

Table 2 Encapsulation efficiency, encapsulation capacity and yield

values of PEO/-CD complex microparticles a-dMean with different subscripts in the same column are significantly different (P <0.05). The data presented tree times

experiment. Table 2shows the results of encapsulation efficiency and capacity and the yield value of PEO/-CD inclusion complexes through the co-precipitation method. The EEs ranged from 8.0940g/100g to 72.8046g/100g and the ECs ranged from 2.4570 g/100g to 35.3649 g/100g, decreasing as the initial essential oil content increased. This finding suggests that the saturation capacity of -CD to EO occurred at a lower of EO content. As a result, the PEO/-CDinclusion complexes microparticles were oily and sticky due to a lack of encapsulation in -CD cavity. In the other hand, the yield value were not significantly different (P < 0.05) from each ratio. The inefficiency of PEO/-CD inclusion complexes at high EO loadingwere due to the insufficient of material and coating. These effecting factors were due to insufficient quantities of-CD,having exceeding quantities of EO, the cavity size of -CD,and molecular structure of EO. A previous study stated that the maincomponent of PEO were eugenol, chavicol, chavibitoland carvacrol(Wuttithamawech, 1997)which were all classified as alcohol group with benzene ring. Therefore, these largemolecules were barriers in accessingthe -CD cavity. C.Effect of antimicrobial activity of BEO and PEO/-CDinclusion complexes According to the EE, EC and yield results (Table 1 and 2), the BEO and PEO/-CD inclusion complexes prepared by a co-precipitation method at 50:50 ratio and 25: 75 ratio respectively showed the highest (P < 0.05) encapsulation. Thus, these ratios were selected to be tested for their antimicrobial activities. The minimum inhibitory concentration (MIC) and the minimum bactericidal concentration (MBC) ofBEOand PEO/-CD inclusion complexes on the six strain of



54

bacteria areshown in Tables 3 and 4 respectively. The

Table 3 Minimum inhibitory and bactericidal concentration(MIC,

MBC) against six strains of bacteria for BEO/-CD inclusion complexes

The value base on the actual concentration of BEO/-CD inclusion complexes. MIC values for BEO/-CDencapsulated ranged from 0.3-10 %w/v (Table 3)while the MIC values for PEO/-CD encapsulated ranged from 0.15-5 %w/v(Table 4)respectively.Considering the MBC values, it can be observed that MBC values for BEO/-CDinclusion complexes were 10 %w/v of all six microorganismswhile the PEO/-CD inclusion complexes ranged from 1.25 -10 %w/v as shown in Table 3 and 4 respectively. From the results on table 3, we found thatBEO/-CD inclusion complexes showed inhibitory effects against E. coli 6795 and S. Typhimurium ATCC 13311with the lowest MICs (0.3 %w/v). These results indicated that BEO/-CD inclusion complexesis highly effective against Gram-negative bacteria except forP. aeruginosa ATCC 27853 due to its antibioticresistancebacteria property of P. aeruginosa( Lambert, 2002).While,the bactericidal capacity were 10 %w/v of the sixtested bacteria.The high MBC value correspondsto previous studiesof loading trans-cinnamaldehyde-CD complex microparticles without being fully releasedfrom the inclusion complex (Heges, et al., 1995).Thisobservation suggests that control release of EO occurred in BEO/-CD inclusion complexes. In table 4, PEO/-CD inclusion complexes displayed inhibitory effects against B. cereus ATCC 11778,S. aureus ATCC 25923, and Y. enterocolitica DMST 17911 with the lowest concentration at 0.15 %w/v and is also highly effective against E. coli 6795 and S. Typhimurium ATCC 13311 with 0.3 %w/v concentration.In considering the MBC value, we foundthatPEO/-CD inclusion complexes showedhigh bactericidal capacity on four bacterias tested anddisplayed bactericidal capacity onY. enterocoliticaDMST 17911(10 %w/v), but it was not effectiveagainst P. aeruginosa ATCC 27853. The high inhibitory and bactericidal effect ofPEO/-CD inclusion complexescorresponds with the previous studiesthat reportscyclodextrin and their complexes can form large water

soluble aggregates in aqueous solution and are able tosolubilize lipophilic drug (Choi et al, 2009; Loftson etal, 2004). Similar results were found in other studiesthatduring CD/EO complexation, each EO molecule isindividually surrounded by a CD, (Samperio, et al., 2010)

Table 4 Minimum inhibitory and bactericidal concentration(MIC,

MBC) against six strains of bacteria for PEO/-CDinclusion complexes

The value base on the actual concentration of PEO/-CD inclusion complexes. leading to advantageous changes in the chemical and physical properties of gest molecules. It has been suggested that CDs are able to form inclusion complexes through individual molecular interaction as well as more complicated and largermolecular structure that can increase the solubility of hydrophobic compounds(Loftson, et al., 2003).Additionally, other studies suggested that EO/-CD–derived compounds hasan effect on the lipid membrane componentand can causecomplete deterioration of cellular membranes at significantly increased concentration(Liang, at al., 2012).Therefore, PEO/-CD inclusion complexesis effective on tested microorganisms.When the two essential oils encapsulated in -CDcomplexes were compared it was found thatPEO/-CD inclusion complex is more effective than the BEO/-CD inclusion complex although it has lower %EE and %EC due to differencein the main componentof each essential oil. CONCLUSIONS In this study, BEO and PEO-CD complexes microparticles was successful by co-precipitation method. The present study demonstrated initial essential oil content and type of essential oil were effect on physico-chemical and their antimicrobial properties. The EEs and ECs of BEO/-CDcomplexincreased as the BEO content increased, but the yield value declined. On the other hand, the EEs and ECs of PEO -CD complexes increased with decreasing PEO content, while yield values were not significantly different (P<0.05).When the antimicrobial properties of BEO and PEO microparticles were compared found that PEO is more effective than the BEO microparticles. The analyzation of BEO and PEO inclusion complexesby SEM revealed successful encapsulationin the -CD molecules. The images showed irregular shapes and sharp



55

edges.Furthermore, BEO and PEO inclusion complexes showed inhibitory and bactericidal effects against the six tested microorganisms. This study suggested that BEO and PEO inclusion complexes could be applied to natural food preservative in food industries. ACKNOWLEDGEMENTS The authors acknowledge financial support from Nation Research Council of Thailand, 2017 and Thammasat University. And I would like to thank Thai-China Flavours and Fragrances Industry Co.,Ltd. in supporting thematerial in this study. REFERENCES [1] Angioni, A. Barra, E. Cereti, D. Barile, J. D. Coïsson, M.Arlorio, S.

Dessi, V. Coroneo, and P. Cabras, “Chemical composition, plant genetic differences, antimicrobial and antifungal activity investigation of the essential oil of Rosmarinusofficinalis L.” Journal of Agriculture and Food Chemistry, Vol. 52, pp. 3530-3535, 2005.

[2] Raman and S. C. Kang, “In vitro control of food-borne and food spoilage bacteria by essential oil and ethanol extract of Lonicera japonica Thumb” Food Chemistry, Vol. 116, pp. 670-675, 2009.

[3] S. Manso, F. Cacho-nerín, R. Becerril and C. Nerín, “Combined analytical and microbiological tools to study the effect on Aspergillusflavus of cinnamon essential oil contained in food packaging” Food control, Vol. 30, pp. 370-378, 2013

[4] P. Sirirugsa, “ Studying of Zingiberaceae in Thailand” Journal of Science, Naresuan University, Vol. 5, no.2 , pp.119-128, 2008.

[5] Chaveerach, T. Tanee, R. Sudmoon and P. Sudmool, “The Genus Piper in Thailand ” 2nd Edition , KhonKaen:KhonKaen Printing Co., Ltd. factory.

[6] J. Wang, C. Yanping, S. Baoguo and W. Chengtao, “Physicochemical and release characterization of garlic oil- ß-cyclodextrin inclusion complexes” Journal of Food Chemistry, Vol. 127, pp. 1680-1685, 2011.

[7] E. L. Hill., C. Gomes, and T. M. Taylor, “Characterization of beta-cyclodextrin inclusion complexes containing essential oils (trans-cinnamaldehyde, eugenol, cinnamon bark and clove bud extracts) for antimicrobial delivery applications” Journal of Food Science and Technology, Vol.51, pp. 86-93, 2013.

[8] S. Burt, “Essential oil: Their antimicrobial properties and potential applications in foods-a review” International Journal of Food Microbiology, Vol. 94, pp. 223-253, 2004.

[9] F. Bakkali, S. Averbeck, D. Averbeck and M. Idaomar, “Biological effects of essential oils – A review ” Food and Chemical Toxicology, Vol. 46, pp. 446-475, 2008.

[10] U. Klinkesorn, P. Sophanodora, P. Chinachoti, E. A., Deker and D. J. McClements, “Characterization of spray-dried tuna oil emulsified in two-layered interfacial membranes prepared using electrostatic layer-by-layer deposition” Food Research International, Vol. 39, pp. 449-457, 2006.

[11] Padukka, B. Bhandari, and B. D'Arcy, “Evaluation of various extraction methods of encapsulated oil from β -cyclodextrin-lemon oil complex powder” Journal of Food Composition and Analysis. Vol.13, pp. 59-70, 2000.

[12] E. J. Seo, S. G. Min and M. J. Choi, “Release characteristics of freeze-dried eugenol encapsulated with b-cyclodextrin by molecular inclusion method” Journal of Microencapsulation, Vol. 27, pp. 496-505, 2010.

[13] Dima, M. Cotarlet, B. Tiberius, G. Bahrim, P. Alexe and S. Dima, “Encapsulation of Coriander Essential Oil in Beta-Cyclodextrin: Antioxidant and Antimicrobial Properties Evaluation” Romanian Biotechnological Letters, Vol. 19, no. 2, 2014.

[14] M. J. Choi, U. Ruktanonchai, S. G. Min, J. Y. Chun and A. Soottitantawat, “Physical characteristics of fish oil encapsulated by β–cyclodextrin using an aggregation method or polycaprolactone using an emulsion-diffusion method” Food Chemistry, Vol. 119, pp. 1694-1703, 2010.

[15] C. Rattanakreetakul, S. Jamkratoak, C. Leksomboon, N.Farungsang and U. Farungsang, “Chemical Properties of Essential Oil from Krachai and Their Antifungus Plant Pathogens” In The Proceeding of 43th Kasetsart University Annual Conference. Kasetsart University, Bangkok, Thailand, 2005.

[16] W. Wuttithamawech, (1997). Thai Herb Encyclopedia “Thai Pharmaceutical Principles” O. S. Printing House. Bangkok, Thailand.

[17] P. A. Lambert, “Mechanisms of antibiotic resistance in Pseudomonas aeruginosa” Journal of the Royal Society of Medicine, Vol. 95, pp. 22-26, 2002.

[18] R. Hedges, W. J. Shieh and C. T. Sikorski. (1995). Use of cyclodextrins for encapsulation in the use and treatment of food products. In S. J. Risch& G. A. Reineccius (Eds.), Encapsulation and controlled release of food ingredients, Vol. 590, (pp. 60-73). Washington, DC: American Chemical Society.

[19] T. Loftsson, M. Masson and M. E. Brewster, “Self-association of cyclodextrins and cyclodextrin complexes” Journal of Pharmaceutical Sciences, Vol. 93. pp. 1091-1099, 2004.

[20] Samperio, R. Boyer, W. N. Eigel, K.W. Holland, J. S. McKinney, S. F. O’Keefe, R. Smith and J. E. Macry, “Enhanment of plant essential oils’ aqueous solubility and stability using alpha and beta cyclodextrin” Journal of Agriculture and Food Chemistry, Vol. 58, pp.12950-12956, 2010.

[21] T. Loftson, K. Matthiasson and M. Masson, “The effects of organic salts on the cyclodextrinsolubilization of drugs” International Journal of Pharmaceutics, Vol. 262, pp.101-107, 2003.

[22] H. Liang, Y. Qupeng, V. Frank and L. Fei, “Effect ofcyclodextrins on the antimicrobial activity of plant-derived essential oil compounds” Journal of Food Chemistry, Vol.135, pp.1020-1027, 2012.

PROCEEDINGS OF€¦ · filmmakers, and feminist filmmakers can be nurtured through the festival. At the same time, the festival increases the gender self-awareness of the women who

Documents