Top Banner

of 226

IJACSA_Volume3 No11

Apr 03, 2018



Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
  • 7/28/2019 IJACSA_Volume3 No11


  • 7/28/2019 IJACSA_Volume3 No11



    www.thesa i .o rg | in fo@thesa i .o rg

  • 7/28/2019 IJACSA_Volume3 No11


    (IJACSA) International Journal of Advanced Computer Science and Applications,Vol. 3, No.11, 2012


    Editorial Preface

    From the D esk of M anaging Editor IJACSA seems to have a cult following and was a humungous success during 2011. We at The Science and Information

    Organization are pleased to present the November 2012 Issue of IJACSA.

    While it took the radio 38 years and the television a short 13 years, it took the World Wide Web only 4 years to reach 50

    million users. This shows the richness of the pace at which the computer science moves. As 2012 progresses, we seem to

    be set for the rapid and intricate ramifications of new technology advancements.

    With this issue we wish to reach out to a much larger number with an expectation that more and more researchers get

    interested in our mission of sharing wisdom. The Organization is committed to introduce to the research audience

    exactly what they are looking for and that is unique and novel. Guided by this mission, we continuously look for ways to

    collaborate with other educational institutions worldwide.

    Well, as Steve Jobs once said, Innovation has nothing to do with how many R&D dollars you have, its about the people

    you have. At IJACSA we believe in spreading the subject knowledge with effectiveness in all classes of audience.

    Nevertheless, the promise of increased engagement requires that we consider how this might be accomplished,delivering up-to-date and authoritative coverage of advanced computer science and applications.

    Throughout our archives, new ideas and technologies have been welcomed, carefully critiqued, and discarded or

    accepted by qualified reviewers and associate editors. Our efforts to improve the quality of the articles published and

    expand their reach to the interested audience will continue, and these efforts will require critical minds and careful

    consideration to assess the quality, relevance, and readability of individual articles.

    To summarise, the journal has offered its readership thought provoking theoretical, philosophical, and empirical ideas

    from some of the finest minds worldwide. We thank all our readers for their continued support and goodwill for IJACSA.

    We will keep you posted on updates about the new programmes launched in collaboration.

    We would like to remind you that the success of our journal depends directly on the number of quality articles submitted

    for review. Accordingly, we would like to request your participation by submitting quality manuscripts for review andencouraging your colleagues to submit quality manuscripts for review. One of the great benefits we can provide to our

    prospective authors is the mentoring nature of our review process. IJACSA provides authors with high quality, helpful

    reviews that are shaped to assist authors in improving their manuscripts.

    We regularly conduct surveys and receive extensive feedback which we take very seriously. We beseech valuable

    suggestions of all our readers for improving our publication.

    Thank you for Sharing Wisdom!

    Managing EditorIJACSAVolume 3 Issue 11November2012ISSN 2156-5570(Online)ISSN 2158-107X (Print)

    2012 The Science and Information (SAI) Organization

  • 7/28/2019 IJACSA_Volume3 No11


    (IJACSA) International Journal of Advanced Computer Science and Applications,Vol. 3, No.11, 2012


    Associate Editors

    Dr.Zuqing Zhu

    Service Provider Technology Group of Cisco Systems, San Jose

    Domain of Research: Research and development of wideband access routers for hybridfibre-coaxial (HFC) cable networks and passive optical networks (PON)

    Dr.KaLok Man

    Department of Computer Science and Software Engineering at the Xi'an Jiaotong-

    Liverpool University, China

    Domain of Research: Design, analysis and tools for integrated circuits and systems;

    formal methods; process algebras; real-time, hybrid systems and physical cyber

    systems; communication and wireless sensor networks.

    Dr.SasanAdibiTechnical Staff Member of Advanced Research, Research In Motion (RIM), Canada

    Domain of Research: Security of wireless systems, Quality of Service (QoS), Ad-Hoc

    Networks, e-Health and m-Health (Mobile Health)


    Associate Professor in the Department of Computer Science at the University of West


    Domain of Research: Database and Data Mining.

    Dr. T. V. Prasad

    Dean, Lingaya's University, India

    Domain of Research: Bioinformatics, Natural Language Processing, Image Processing,

    Expert Systems, Robotics

    Dr.Bremananth R

    Research Fellow, Nanyang Technological University, Singapore

    Domain of Research: Acoustic Holography, Pattern Recognition, Computer Vision, Image

    Processing, Biometrics, Multimedia and Soft Computing

  • 7/28/2019 IJACSA_Volume3 No11


    (IJACSA) International Journal of Advanced Computer Science and Applications,Vol. 3, No.11, 2012


    Reviewer Board Members

    A KathirvelKarpaga Vinayaka College of Engineering and

    Technology, India

    A.V. Senthil KumarHindusthan College of Arts and Science

    Abbas KarimiI.A.U_Arak Branch (Faculty Member) & Universiti

    Putra Malaysia

    Abdel-Hameed A. BadawyUniversity of Maryland

    Abdul WahidGautam Buddha University

    Abdul HannanVivekanand College

    Abdul Khader Jilani SaudagarAl-Imam Muhammad Ibn Saud Islamic University

    Abdur Rashid KhanGomal Unversity

    Aderemi A. AtayeroCovenant University

    Ahmed Boutejdar Dr. Ahmed Nabih Zaki Rashed

    Menoufia University, Egypt

    Ajantha HerathUniversity of Fiji

    Ahmed Sabah AL-JumailiAhlia University

    Akbar Hossain Albert Alexander

    Kongu Engineering College,India

    Prof. Alcinia Zita SampaioTechnical University of Lisbon

    Amit VermaRayat & Bahra Engineering College, India

    Ammar Mohammed AmmarDepartment of Computer Science, University of


    Anand NayyarKCL Institute of Management and Technology,Jalandhar

    Anirban SarkarNational Institute of Technology, Durgapur, India

    Arash Habibi LashakriUniversity Technology Malaysia (UTM), Malaysia

    Aris SkanderConstantine University

    Ashraf Mohammed Iqbal

    Dalhousie University and Capital Health

    Asoke NathSt. Xaviers College, India

    Aung Kyaw OoDefence Services Academy

    B R SARATH KUMARLenora College of Engineering, India

    Babatunde Opeoluwa AkinkunmiUniversity of Ibadan

    Badre BossoufiUniversity of Liege

    Balakrushna TripathyVIT University

    Basil HamedIslamic University of Gaza

    Bharat Bhushan AgarwalI.F.T.M.UNIVERSITY

    Bharti Waman GawaliDepartment of Computer Science &


    Bremananth RamachandranSchool of EEE, Nanyang Technological University

    Brij GuptaUniversity of New Brunswick

    Dr.C.Suresh Gnana DhasPark College of Engineering and Technology,India

    Mr. Chakresh kumarManav Rachna International University, India

    Chandra Mouli P.V.S.S.RVIT University, India

    Chandrashekhar MeshramChhattisgarh Swami Vivekananda Technical


    Chao Wang Chi-Hua Chen

    National Chiao-Tung University

    Constantin POPESCUDepartment of Mathematics and Computer

    Science, University of Oradea

    Prof. D. S. R. MurthySNIST, India.

    Dana PETCUWest University of Timisoara

    David GreenhalghUniversity of Strathclyde
  • 7/28/2019 IJACSA_Volume3 No11


    (IJACSA) International Journal of Advanced Computer Science and Applications,Vol. 3, No.11, 2012


    Deepak GargThapar University.

    Prof. Dhananjay R.KalbandeSardar Patel Institute of Technology, India

    Dhirendra MishraSVKM's NMIMS University, India

    Divya Prakash ShrivastavaEL JABAL AL GARBI UNIVERSITY, ZAWIA

    Dr.Dhananjay Kalbande Dragana Becejski-Vujaklija

    University of Belgrade, Faculty of organizational


    Driss EL OUADGHIRI Firkhan Ali Hamid Ali


    Fokrul Alom MazarbhuiyaKing Khalid University

    Frank IbikunleCovenant University

    Fu-Chien KaoDa-Y eh University

    G. SreedharRashtriya Sanskrit University

    Gaurav KumarManav Bharti University, Solan Himachal


    Ghalem BelalemUniversity of Oran (Es Senia)

    Gufran Ahmad AnsariQassim University

    Hadj Hamma TadjineIAV GmbH

    Hanumanthappa.JUniversity of Mangalore, India

    Hesham G. IbrahimChemical Engineering Department, Al-Mergheb

    University, Al-Khoms City

    Dr. Himanshu AggarwalPunjabi University, India

    Huda K. AL-JoboriAhlia University

    Iwan SetyawanSatya Wacana Christian University

    Dr. Jamaiah Haji YahayaNorthern University of Malaysia (UUM), Malaysia

    Jasvir SinghCommunication Signal Processing Research Lab

    Jatinderkumar R. SainiS.P.College of Engineering, Gujarat

    Prof. Joe-Sam ChouNanhua University, Taiwan

    Dr. Juan Jos Martnez CastilloYacambu University, Venezuela

    Dr. Jui-Pin YangShih Chien University, Taiwan

    Jyoti Chaudharyhigh performance computing research lab

    K RamaniK.S.Rangasamy College of Technology,

    Tiruchengode K V.L.N.Acharyulu

    Bapatla Engineering college


    Ka Lok ManXian Jiaotong-Liverpool University (XJTLU)

    Dr. Kamal ShahSt. Francis Institute of Technology, India


    Kashif NisarUniversiti Utara Malaysia

    Kavya Naveen Kayhan Zrar Ghafoor

    University Technology Malaysia

    Kodge B. G.S. V. College, India

    Kohei AraiSaga University

    Kunal PatelIngenuity Systems, USA

    Labib Francis GergisMisr Academy for Engineering and Technology

    Lai Khin WeeTechnischen Universitt Ilmenau, Germany

    Latha ParthibanSSN College of Engineering, Kalavakkam

    Lazar StosicCollege for professional studies educators,


    Mr. Lijian SunChinese Academy of Surveying and Mapping,


    Long ChenQualcomm Incorporated

    M.V.RaghavendraSwathi Institute of Technology & Sciences, India.

    M. Tariq BandayUniversity of Kashmir

    Madjid Khalilian
  • 7/28/2019 IJACSA_Volume3 No11


    (IJACSA) International Journal of Advanced Computer Science and Applications,Vol. 3, No.11, 2012


    Islamic Azad University

    Mahesh ChandraB.I.T, India

    Mahmoud M. A. Abd EllatifMansoura University

    Manas deepMasters in Cyber Law & Information Security

    Manpreet Singh MannaSLIET University, Govt. of India

    Manuj DarbariBBD University

    Marcellin Julius NKENLIFACKUniversity of Dschang

    Md. Masud RanaKhunla University of Engineering & Technology,


    Md. Zia Ur RahmanNarasaraopeta Engg. College, Narasaraopeta

    Messaouda AZZOUZIZiane AChour University of Djelfa

    Dr. Michael WattsUniversity of Adelaide, Australia

    Milena BogdanovicUniversity of Nis, Teacher Training Faculty in


    Miroslav BacaUniversity of Zagreb, Faculty of organization and

    informatics / Center for biomet

    Mohamed Ali MahjoubPreparatory Institute of Engineer of Monastir

    Mohammad TalibUniversity of Botswana, Gaborone

    Mohamed El-Sayed Mohammad Yamin Mohammad Ali Badamchizadeh

    University of Tabriz

    Mohammed Ali HussainSri Sai Madhavi Institute of Science &


    Mohd Helmy Abd WahabUniversiti Tun Hussein Onn Malaysia

    Mohd Nazri IsmailUniversity of Kuala Lumpur (UniKL)

    Mona ElshinawyHoward University

    Monji KherallahUniversity of Sfax

    Mourad AmadLaboratory LAMOS, Bejaia University

    Mueen UddinUniversiti Teknologi Malaysia UTM

    Dr. Murugesan NGovernment Arts College (Autonomous), India

    N Ch.Sriman Narayana IyengarVIT University

    Natarajan SubramanyamPES Institute of Technology

    Neeraj BhargavaMDS University

    Nitin S. ChoubeyMukesh Patel School of Technology

    Management & Eng

    Noura AkninAbdelamlek Essaadi

    Om Sangwan Pankaj Gupta

    Microsoft Corporation

    Paresh V VirpariaSardar Patel University

    Dr. Poonam GargInstitute of Management Technology,



    Pradip JawandhiyaJawaharlal Darda Institute of Engineering &


    Rachid SaadaneEE departement EHTP

    Raghuraj Singh

    Raj Gaurang TiwariAZAD Institute of Engineering and Technology

    Rajesh KumarNational University of Singapore

    Rajesh K ShuklaSagar Institute of Research & Technology-

    Excellence, India

    Dr. Rajiv DharaskarGH Raisoni College of Engineering, India

    Prof. Rakesh. LVijetha Institute of Technology, India

    Prof. Rashid SheikhAcropolis Institute of Technology and Research,India

    Ravi PrakashUniversity of Mumbai

    Reshmy KrishnanMuscat College affiliated to stirling University.U

    Rongrong JiColumbia University

    Ronny Mardiyanto
  • 7/28/2019 IJACSA_Volume3 No11


    (IJACSA) International Journal of Advanced Computer Science and Applications,Vol. 3, No.11, 2012


    Institut Teknologi Sepuluh Nopember

    Ruchika MalhotraDelhi Technological University

    Sachin Kumar AgrawalUniversity of Limerick

    Dr.Sagarmay DebUniversity Lecturer, Central Queensland

    University, Australia Said Ghoniemy

    Taif University

    Saleh Ali K. AlOmariUniversiti Sains Malaysia

    Samarjeet BorahDept. of CSE, Sikkim Manipal University

    Dr. Sana'a Wafa Al-SayeghUniversity College of Applied Sciences UCAS-


    Santosh KumarGraphic Era University, India

    Sasan AdibiResearch In Motion (RIM)

    Saurabh PalVBS Purvanchal University, Jaunpur

    Saurabh DuttaDr. B. C. Roy Engineering College, Durgapur

    Sebastian Marius RosuSpecial Telecommunications Service

    Sergio Andre FerreiraPortuguese Catholic University

    Seyed Hamidreza Mohades KasaeiUniversity of Isfahan

    Shahanawaj AhamadThe University of Al-Kharj

    Shaidah JusohUniversity of West Florida

    Shriram Vasudevan Sikha Bagui

    Zarqa University


    Slim BEN SAOUD Dr. Smita Rajpal

    ITM University

    Suhas J ManangiMicrosoft

    SUKUMAR SENTHILKUMARUniversiti Sains Malaysia

    Sumazly SulaimanInstitute of Space Science (ANGKASA), Universiti

    Kebangsaan Malaysia

    Sumit Goyal

    Sunil TanejaSmt. Aruna Asaf Ali Government Post Graduate

    College, India

    Dr. Suresh SankaranarayananUniversity of West Indies, Kingston, Jamaica

    T C. ManjunathHKBK College of Engg

    T C.ManjunathVisvesvaraya Tech. University

    T V Narayana RaoHyderabad Institute of Technology and


    T. V. PrasadLingaya's University

    Taiwo AyodeleLingaya's University

    Tarek Gharib Totok R. Biyanto

    Infonetmedia/University of Portsmouth

    Varun KumarInstitute of Technology and Management, India

    Vellanki Uma Kanta SastrySreeNidhi Institute of Science and Technology

    (SNIST), Hyderabad, India.

    Venkatesh Jaganathan Vijay Harishchandra Vinayak Bairagi

    Sinhgad Academy of engineering, India

    Vishal BhatnagarAIACT&R, Govt. of NCT of Delhi

    Vitus S.W. LamThe University of Hong Kong

    Vuda SreenivasaraoSt.Marys college of Engineering & Technology,

    Hyderabad, India

    Wei Wei Wichian Sittiprapaporn

    Mahasarakham University

    Xiaojing XiangAT&T Labs

    Y SrinivasGITAM University

    Yilun ShangUniversity of Texas at San Antonio

    Mr.Zhao ZhangCity University of Hong Kong, Kowloon, Hong


    Zhixin ChenILX Lightwave Corporation

    Zuqing ZhuUniversity of Science and Technology of China
  • 7/28/2019 IJACSA_Volume3 No11


    (IJACSA) International Journal of Advanced Computer Science and Applications,Vol. 3, No.11, 2012



    Paper 1: Predicting Garden Path Sentences Based on Natural Language Understanding System

    Authors: DU Jia-li, YU Ping-fang

    PAGE 16

    Paper 2: Extending UML for trajectory data warehouses conceptual modelling

    Authors: Wided Ouesatli, Jalel Akaichi

    PAGE 712

    Paper 3: Optimal itinerary planning for mobile multiple agents in WSN

    Authors: Mostefa BENDJIMA,Mohamed FEHAMPAGE 1319

    Paper 4: Genetic procedure for the Single Straddle Carrier Routing ProblemAuthors: Khaled MILI, Faissal MILI

    PAGE 20


    Paper 5: The Virtual Enterprise Network based on IPSec VPN Solutions and Management

    Authors:Sebastian Marius Rosu, Marius Marian Popescu, George Dragoi, Ioana Raluca GuicaPAGE 2634

    Paper 6: Analyzing the Efficiency of Text-to-Image Encryption Algorithm

    Authors:Ahmad Abusukhon, Mohammad Talib,Maher A. NabulsiPAGE 3538

    Paper 7: DNA Sequence Representation and Comparison Based on Quaternion Number System

    Authors:Hsuan-T. Chang, Chung J. Kuo, Neng-Wen Lo, Wei-Z.LvPAGE 3946

    Paper 8: Evaluation of Regressive Analysis Based Sea Surface Temperature Estimation Accuracy with NCEP/GDAS


    Authors:Kohei AraiPAGE 4752

    Paper 9:The Modelling Process of a Paper Folding Problem in GeoGebra 3DAuthors:Muharrem Aktumen, Bekir Kursat Doruk, Tolga Kabaca

    PAGE 5357

    Paper 10:Sensitivity Analysis of Fourier Transformation Spectrometer: FTS against Observation Noise onRetrievals of Carbon Dioxide and Methane

    Authors: Kohei Arai, Hiroshi Okumura,Takuya Fukamachi, Shuji Kawakami,Hirofumi OhyamaiPAGE 5864

  • 7/28/2019 IJACSA_Volume3 No11


    (IJACSA) International Journal of Advanced Computer Science and Applications,Vol. 3, No.11, 2012


    Paper 11: Sensitivity Analysis for Water Vapour Profile Estimation with Infrared: IR Sounder Data Based on


    Authors: Kohei Arai

    PAGE 6570

    Paper 12: Wavelet Based Change Detection for Four Dimensional Assimilation Data in Space and Time DomainsAuthors:KoheiArai

    PAGE 7175

    Paper 13: Method for Image Source Separation by Means of Independent Component Analysis: ICA, Maximum

    Entory Method: MEM, and Wavelet Based Method: WBMAuthors:Kohei Arai

    PAGE 7681

    Paper 14: Multifinger Feature Level Fusion Based Fingerprint Identification

    Authors:Praveen N, Tessamma ThomasPAGE 8288

    Paper 15: Gender Effect Canonicalization for Bangla ASR

    Authors:B.K.M. Mizanur Rahman, Bulbul Ahamed, Md. Asfak-Ur-Rahman, Khaled Mahmud, MohammadNurul Huda

    PAGE 8994

    Paper 16: Three Layer Hierarchical Model for Chord

    Authors: Waqas A. Imtiaz, Shimul Shil, A.K.M Mahfuzur Rahman

    PAGE 9599

    Paper 17: Automatic Facial Expression Recognition Based on Hybrid Approach

    Authors: Ali K. K. Bermani, Atef Z. Ghalwash, Aliaa A. A. YoussifPAGE 100105Paper 18: Design of A high performance low-power consumption discrete time Second order Sigma-Delta

    modulator used for Analog to Digital Converter

    Authors: Radwene LAAJIMI, Mohamed MASMOUDI

    PAGE 106112

    Paper 19: Short Answer Grading Using String Similarity and Corpus-Based Similarity

    Authors:Wael H. Gomaa, Aly A. FahmyPAGE 113119

    Paper 20: Hybrid intelligent system for Sale Forecasting using Delphi and adaptive Fuzzy Back-Propagation Neural


    Authors: Attariuas Hicham, Bouhorma Mohammed, Sofi Anas

    PAGE 120128

  • 7/28/2019 IJACSA_Volume3 No11


    (IJACSA) International Journal of Advanced Computer Science and Applications,Vol. 3, No.11, 2012


    Paper 21: Continuous Bangla Speech Segmentation using Short-term Speech Features Extraction Approaches

    Authors: Md. Mijanur Rahman, Md. Al-Amin Bhuiyan

    PAGE 129137

    Paper 22: A Randomized Fully Polynomial-time Approximation Scheme for Weighted Perfect Matching in the


    Authors: Yasser M. Abd El-Latif, Salwa M. Ali, Hanaa A.E. Essa, Soheir M. Khamis

    PAGE 138143

    Paper 23: Cost Analysis of Algorithm Based Billboard Manger Based Handover Method in LEO satellite Networks

    Authors: Suman Kumar Sikdar, Soumaya Das,Debabrata SarddarPAGE 144150

    Paper 24: An agent based approach for simulating complex systems with spatial dynamics application in the land

    use planning

    Authors: Fatimazahra BARRAMOU, Malika ADDOUPAGE 151157

    Paper 25: Improving Web Page Prediction Using Default Rule Selection

    Authors: Thanakorn Pamutha, Chom KimpanSiriporn Chimplee, Parinya SanguansatPAGE 158163

    Paper 26: Multilayer Neural Networks and Nearest Neighbour Classifier Performances for Image Annotation

    Authors: Mustapha OUJAOURA, Brahim MINAOUI, Mohammed FAKIR

    PAGE 164170

    Paper 27: Minimization of Call Blocking Probability using Mobile Node velocity

    Authors: Suman Kumar Sikdar, Uttam Kumar Kundu, Debabrata Sarddar

    PAGE 171


    Paper 28: A Pheromone Based Model for Ant Based Clustering

    Authors: Saroj Bala, S. I. Ahson, R. P. AgarwalPAGE 180183

    Paper 29: Modern and Digitalized USB Device with Extendable Memory Capacity

    Authors: J.Nandini Meeraa, S.Devi Abirami, N.Indhuja, R.Aravind, C.Chithiraikkayalvizhi, K.Rathina Kumar

    PAGE 184188

    Paper 30: Enhanced Modified Security Framework for Nigeria Cashless E-payment System

    Authors: Fidelis C. Obodoeze , Francis A. OkoyePAGE 189196

    Paper 31: Cashless Society: An Implication To Economic Growth & Development In Nigeria


    PAGE 197203

  • 7/28/2019 IJACSA_Volume3 No11


    (IJACSA) International Journal of Advanced Computer Science and Applications,Vol. 3, No.11, 2012


    Paper 32: The Analysis of the Components of Project-Based Learning on Social Network

    Authors:Kuntida Thamwipat, Napassawan YookongPAGE 204209

    Paper 33: A Database Creation for Storing Electronic Documents and Research of the Staff

    Authors:Pornpapatsorn Princhankol, Siriwan PhacharakreangchaiPAGE 210214

  • 7/28/2019 IJACSA_Volume3 No11


    (IJACSA) International Journal of Advanced Computer Science and Applications,Vol. 3, No. 11, 2012

    1 | P a g e

    Predicting Garden Path Sentences Based on Natural

    Language Understanding System

    DU Jia-li

    School of Foreign Languages/School of LiteratureLudong University/Communication University of China

    Yantai/ Beijing, China

    YU Ping-fang

    School of Liberal Arts/Institute of LinguisticsLudong University/Chinese Academy of Social Sciences

    Yantai/ Beijing, China

    Abstract Natural language understanding (NLU) focusing onmachine reading comprehension is a branch ofnatural language

    processing (NLP). The domain of the developing NLU system

    covers from sentence decoding to text understanding and the

    automatic decoding of GP sentence belongs to the domain of NLU

    system. GP sentence is a special linguistic phenomenon in which

    processing breakdown and backtracking are two key features. If

    the syntax-based system can present the special features of GP

    sentence and decode GP sentence completely and perfectly, NLU

    system can improve the effectiveness and develop theunderstanding skill greatly. On the one hand, by means of

    showing Octav Popescus model of NLU system, we argue that

    the emphasis on the integration of syntactic, semantic and

    cognitive backgrounds in system is necessary. On the other hand,

    we focus on the programming skill of IF-THEN-ELSE statement

    used in N-S flowchart and highlight the function of context free

    grammar (CFG) created to decode GP sentence. On the basis of

    example-based analysis, we reach the conclusion that syntax-

    based machine comprehension is technically feasible and

    semantically acceptable, and that N-S flowchart and CFG can

    help NLU system present the decoding procedure of GP sentence

    successfully. In short, syntax-based NLU system can bring a

    deeper understanding of GP sentence and thus paves the way for

    further development of syntax-based natural language processing

    and artificial intelligence.

    Keywords- Natural language understanding; N-S flowchart;computational linguistics; context free grammar; garden pathsentences.

    I. INTRODUCTIONNatural language understanding (NLU), speech

    segmentation, text segmentation, part-of-speech tagging, wordsense disambiguation, syntactic ambiguity, etc. come underthe umbrella term natural language processing (NLP).[1]The development of NLU is briefly traced from the early yearsof machine translation to today's question answering andtranslation systems. [2-3]NLU today deals with machine

    reading comprehension in artificial intelligence (AI) [4]and isapplied to a diverse set of computer applications.[5-6] Itssubject ranges from simple tasks such as short commandsissued to robots, to complex endeavors such as the fullcomprehension of articles or essays. A machine created tounderstand natural language has been one of the dreams of AIever since computers were invented, and this encourages thesystematic and rapid development of NLU. The efficientunderstanding of natural language requires that computerprogram be able to resolve ambiguities at the syntactic level

    and recover that part of the meaning of its individual wordstaken in isolation.[7-8] The satisfaction of this requirementinvolves complex inference from a large database of world-knowledge, and this makes the designer of computer programsfor NLU face the serious difficulty of algorithm processing.[9]The machine comprehension is embedded in the more generalframe of interpersonal communication and is applied to theperson-machine interaction task.[10] The further integrationhas proved appropriate for the design of effective and robustnatural language interfaces to AI systems. Syntactic gardenpath phenomenon is a major source of uncertainty in NLU.Syntax-supported systems attempt to help machine to get adeeper understanding of garden path sentence and the relatedalgorithms deserve special attention in the future of NLUdeveloping. [11] Many hybrid researches contribute to theimprovement of NLU systems. [12] For example, the fact isestablished that abduction rather than deduction is generallyviewed as a promising way to apply reasoning in NLU.[13]The development of NLU can focus on the design of astochastic model topology that is optimally adapted in qualityand complexity to the task model and the available trainingdata.[14] A comparative information-theoretic study is carriedout to show positional letter analyses, n-gram analyses, wordanalyses, empirical semantic correlations between the Greekand English n-grams, and entropy calculations are useful totext processing and compression, speech synthesis andrecognition as well as error detection and correction.[15]

    Garden path (GP) phenomenon is a special linguisticphenomenon which comprises processing breakdown andbacktracking. GP sentence (e.g. The old dog the footsteps ofthe young) is an originally correct sentence which makesreaders grammatical misinterpretation linger until re-decoding has occurred. An incorrect choice in GP sentenceusually is readers' most likely interpretation, leading readersinitially into an improper parse which, however, finally provesto be a dead end. Thus the processing breaks down and

    backtrack to the original status to search the given informationagain for alternative route to the successful decoding. Theautomatic decoding of GP sentence is a challenge for NLUsystems for machine has to have access to grammatical,semantic and cognitive knowledge in order to understandnatural language smoothly as human brains do.

    In this paper, a syntax-based and algorithm-originatedapproach to understanding GP sentence in natural languagedomain is presented. The discussion consists of four sections.Firstly, a logic-based model of NLU system is shown to
  • 7/28/2019 IJACSA_Volume3 No11


    (IJACSA) International Journal of Advanced Computer Science and Applications,Vol. 3, No. 11, 2012

    2 | P a g e

    provide an overview of NLU. Secondly, N-S flowchart is usedto present the special feature of processing breakdown of GPsentence. Thirdly, context-free grammar (CFG) is introducedto analyze the backtracking of GP sentence in detail and themachines automatic decoding procedure of GP sentence isanalyzed. The last section brings the conclusion.


    In the domain of Intelligent Tutoring Systems, the high-precision NLU system is helpful to accurately determine thesemantic content of learners explanations. For example, theNLU system developed in the context of the GeometryCognitive Tutor combines unification-based syntacticprocessing with Description Logic based semantics to achievethe necessary accuracy level. [16] The syntactic and semanticprocessing of natural language is useful for specific decodingproblems, like metonymy resolution and reference resolution.On the basis of linguistic theories and computationaltechnologies, NLU system architecture building the syntacticstructure and offering the semantic interpretation of learnerexplanations is structured by Octav Popescu in 2005. (Fig. 1)

    Figure. 1. Octav Popescus NLU system architecture

    The research is related to practical application of NLUsystem. Popescus architecture comprises two sections: thesyntactic and the semantic processing subsystems, which areinteractive during the NLU processing. The syntacticprocessing subsystem uses an active chart parser while the

    semantic one bases its action on a Description Logic system.The key part is interface module connecting the NLUsubsystem to the tutor itself, functioning asynchronously to theNLU system, taking the input sentence from the tutor, andpassing the checked words to the chart parser. Systemfunctions in real time until the parser finishes and passes theresulting classifications back to the tutor. The chart parser,adopting linguistic knowledge about the target naturallanguage from the unification grammar and the lexicon, is thecenter of NLU system. Words of a sentence are taken one by

    one by the parser according to grammar rules, and featurestructures, which can store lexical, syntactic, and semanticproperties of corresponding words and phrases, are builtsimultaneously.

    Popescus system highlights the integration of syntacticand semantic information during the machine decoding,showing that hybrid knowledge is necessary for machine tounderstand natural language. GP sentence is a more complex

    linguistic phenomenon than the common sentence since GPsentence possesses special features of processing breakdownand backtracking. Therefore, the emphasis of integration oflinguistics and computational science is helpful for system toautomatically decode GP sentence. The computational skill(e.g. N-S flowchart) can present in detail the syntacticprocessing procedure, making learners have access to themachine understanding of GP sentence and thus enhancing theeffectiveness of system.


    GP sentence seems to be a grammatically correct sentenceat the original processing stage and is usually used in syntax,

    linguistics, psycholinguistics, and computational linguistics.According to syntactic theory, while a person reads a GPsentence, he builds up an original meaning structure andprocesses natural language one word at a time. With theadvancement of processing, the person finds that he has beenconstructing an incorrect structure and that the next word orphrase cannot be incorporated into the structure originallycreated thus far. The "garden path" is a reference to the saying"to be led down the garden path", meaning "to be misled".Examples are as follows: The horse raced past the barn fell,The man who hunts ducks out on weekends, The cottonclothing is made of grows in Mississippi, The prime numberfew, Fat people eat accumulates, The old man the boat,The tycoon sold the offshore oil tracts for a lot of money

    wanted to kill JR, etc (Some examples cited are from The verbs (e.g.,raced, ducks, grows, number, accumulates) in theabove examples are not the words which a reader is assumingto be the verbs. Some words that look as though they weremodifying nouns are actually being used as nouns and somewords which look as though they were nouns are verbs. Therevised non-GP sentences are as follows: The horse that wasraced past the barn fell down, The man, who hunts, ducksout on weekends., The cotton, of which clothing is made,grows in Mississippi, The prime (group) number few, Fatthat people eat accumulates, The old (people) man the boat,The tycoon, who was sold the offshore oil tracts for a lot ofmoney, wanted to kill JR, etc. Please see the explanation in

    Fig. 2.

    Fig. 2 shows us that GP sentence which has specialfeatures of processing breakdown and backtracking isexclusive of multi-meanings, namely, only one meaninginvolved in the final result of decoding even though thedeviation of the original meaning and the final meaning isobvious. The special features of GP sentence can be presentedwith the help of computational linguistics skills, e.g. N-Sflowchart which was raised by Nassi and Shneiderman [17] to
  • 7/28/2019 IJACSA_Volume3 No11


    (IJACSA) International Journal of Advanced Computer Science and Applications,Vol. 3, No. 11, 2012

    3 | P a g e

    analyze structured programming [18-19]. This kind of analysisis helpful for the effectiveness of discussion boards. [20]

    Figure. 2. Explanation for garden path sentence

    Flowcharts IF-THEN-ELSE statement in computationalprocessing can be used to show the procedure of decoding GPsentence now that this technique adopts a dichotomy betweenYes and No. For example, Fig.3 consists of four pairs ofdichotomies, namely, A and not A, B and not B, C and not C,D and not D. Nassi and Shneiderman introduce seven possibleprocessing results, i.e. (1) not A; (2) not A and not B; (3) notA and B; (4) A and not C; (5) A and not C and not D; (6) A

    and not C and D; (7) A and C.

    Figure. 3 Nassi and Shneidermans flowchart of IF-THEN-ELSE statement

    The dichotomy of IF-THEN-ELSE statement can beapplied to analysis of GP sentence which possesses processingbreakdown and backtracking.

    Example 1. The new record the song.

    Figure. 4 The example of two pairs of dichotomies of N-S flowchart

    There are two pairs of dichotomies involved in Fig. 4. The

    first pair is used to distinguish the verb definition and the noundefinition of record. If the verb definition is chosen, themachine decoding will run along the left column which is thegrammatically correct presentation.

    Otherwise, the noun definition of record is thealternative choice; then the right column is the preferablechoice. According to the cognitive rules, Det+Adj+Noun isthe prototype model in which Adj is used to modify theNoun. Therefore, the right column is the original decodingprocedure until system fails to parse more. The decoding hasto return to the triangle to choose the verb definition ofrecord till every string involved is processed completely.

    The second pair of dichotomy appears after the

    grammatical requirements are met successfully. That is to say,system usually abides by the grammatical rules firstly for thesyntactic cohesion is the basic condition for decoding GPsentence.

    After satisfying the grammatical requirements, system thenconsiders cognitive cohesion. This is the reason why cognitiverules are put after the grammatical decoding. Example 1 isaccepted by system according to cognitive rules and thereforethe result is output finally. If the sentence The dead recordthe song which has the similar decoding procedure likeExample 1 is submitted to system, the cognitive disagreementmakes system reject the decoding and no final result will beoutput.

    Example 2. The old make the young man the boat.
  • 7/28/2019 IJACSA_Volume3 No11


    (IJACSA) International Journal of Advanced Computer Science and Applications,Vol. 3, No. 11, 2012

    4 | P a g e

    Figure.5 The example of four pairs of dichotomies of N-S flowchart

    Fig. 5 consists of four pairs of dichotomies of N-Sflowchart. The first pair is used to differentiate the verbdefinition from the noun definition of make. If the verbdefinition is chosen, the left column is system s decoding path.Otherwise, the noun definition in the right column is thechoice. The second central triangle in left column is applied todistinguish the verb definition from the noun definition ofman. The result is that the noun one is in left part and theverb one is in right part.

    The third pair in right column is similar decodingprocedure as the second pair. The difference between thesecond and the third lies in make in the former is a verbwhile make in the latter is a noun. The last pair ofdichotomy is used to test the output from the cognitive level. Ifthe result is for the cognitive rules, the left column is activated

    and it will be output successfully. Otherwise, the right columnstarts and the failed system has to return to the first pair ofdichotomy to parse again.

    For example, S1 is a grammatical correct output ofExample 2, and it means that The old people make the youngman be the boat which is against the cognitive rules .Therefore, S1 is rejected by system and has to return to thefirst central triangle until the correct S2 is chosen. S2means The old people make the young people sail the boat.

    From the discussion above, we can find that N-S flowchartis useful to analyze GP sentence since this technique has theobvious function by which the special features of processingbreakdown and backtracking can be presented in detail.

    Besides N-S flowchart, N. Chomskys context-freegrammar (CFG) is often applied to natural languageunderstanding and the formalized processing is helpful indecoding GP sentence.


    A context-free grammar (CFG) is also called a phrasestructure grammar in formal language theory. According to the

    grammar, it can naturally generate a formal language andclauses can be embedded inside clauses arbitrarily deeply. Interms ofproduction rules, every production of a CFG is of theform: Vw. Generally speaking, V is a singlenonterminal symbol, and w is a string of terminals and/ornonterminals. CFG used to analyze the syntax of naturallanguages plays an important role in the description anddesign of programming languages in NLU system.

    Noam Chomsky has argued that natural languages arebased on CFG and the processing procedure can be presentedby formal languages. In Chomskys opinion, if ABCorAa can be accepted by system in which A (B.C)represents the nonterminal symbol and a is regarded asterminal symbol, the CFG (usually called Chomsky NormalForm) is a dichotomy possessing the binary tree form. Thusthe programming languages are designed to decode GPsentence.

    According to formal languages, quaternion parameters areinvolved in the programming, namely, Vn (the collection ofnonterminal symbols), Vt (the collection of terminalsymbols), S (the original start), and P (rewriting

    programs). Thus, the natural language can be understood by amachine if the programming language is created. The analysisof GP sentence of Example 2 is as follows in Fig.6.
  • 7/28/2019 IJACSA_Volume3 No11


    (IJACSA) International Journal of Advanced Computer Science and Applications,Vol. 3, No. 11, 2012

    5 | P a g e

    Figure 6 CFG-based understanding of Example 2

    The whole processing procedure of Example 2 is shownabove and comprises 40 steps for a machine to understand the

    GP sentence. The flowchart of the decoding consists of fourparts. The first part is the production of NP+NP+NP; thesecond, NP+S3; the third, S1; the fourth, S2.

    In the first part, not all strings are decoded completelyaccording to the rewriting rules and correspondingly systemfails to go ahead. The decoding returns to the central trianglewhere man is considered a verb or a noun. The verbdefinition ofman is chosen since the noun definition provesto be a dead end. Thus, system comes to the second part ofprocessing procedure in which NP+S3 is the conclusion.The decoding numbers of first part is 1-2-3-4-5-6-7-8-9-10-11-12-12-11-10-9-8-7.

    The second part is the result of backtracking of

    NP+NP+NP. Man is alternatively regarded as a verb andsystem reads the sentence of The old make the young manthe boat into NP+S3 in which make is considered a noun.Since no more rewriting rules support the systematic decodingof NP+S3, the syntactic output is against the grammaticalrequirements and as a result, system rejects the final output inwhich both the noun definition and the verb definition ofman are chosen as the syntactic choices when make isused as a noun. Therefore, system has to backtrack to anothercentral triangle where make is considered to be a verb,which brings the appearance of the third part. The decodingnumbers of second part is 13(7)-14-15-16-17-18-19-20-20-19-18-17-16-15-14-13-12-11-10-9-8-7-6-5-4-3.

    The third part is the result of backtracking ofNP+NP+NP and NP+S3 where noun definition ofmakeis preferable choice now that, according to cognitive andpsychological prototype theory, Det+Adj+Noun is thedefault setting. Since the noun definition of make results ina dead end, system becomes to considermake a verb. Thusthe left central triangle is activated. If man is chosen as anoun, S1 in which Example 2 means The old people make theyoung man (be) the boat is the temporary decoding result.There is no cognitive cohesion involved in S1 even though it

    is grammatical correct. System rejects S1 as a successfuloutput (see Fig. 5) and backtracks to the central triangle whereman can be chosen as a verb, which results in the decodingof S2. The decoding numbers of third part is 21(3)-22-23-24-25-26-27-28-29-30-31-32-32-31-30-29-28-27-26-25.

    The last part is the result of backtracking ofNP+NP+NP,NP+S3 and S1. The only choice for system is that bothmake and man should be verbs. The meaning of Example

    2 is The old people make the young people sail the boat.This explanation satisfies the grammatical, cognitive andsemantic requirements and is accepted by system perfectly.The automatic understanding of GP sentence in NLU systemis completely concluded. The decoding numbers of last part is32(25)-33-34-35-36-37-38-39-40.

    CFG-based natural language understanding is helpful toanalyze the special features of GP sentence. Both processingbreakdown and backtracking can be presented on the basis ofCFG grammar. For example, in the decoding of Example 2, allthe figures in parentheses refer to the stage in whichprocessing breaks down and the converse decoding numbersrepresent the stage in which backtracking occurs. The whole

    decoding numbers of NLU system is as follows: 1-2-3-4-5-6-7-8-9-10-11-12-12-11-10-9-8-7-13(7)-14-15-16-17-18-19-20-20-19-18-17-16-15-14-13-12-11-10-9-8-7-6-5-4-3-21(3)-22-23-24-25-26-27-28-29-30-31-32-32-31-30-29-28-27-26-25-32(25)-33-34-35-36-37-38-39-40.

    V. CONCLUSIONNatural language understanding (NLU) dealing with

    machine reading comprehension is the umbrella term ofnatural language processing (NLP). The advancement ofmachine technologies and computational skills develops NLUsystem. Its subject ranges from sentence decoding to textunderstanding. The decoding of GP sentence, a speciallinguistic phenomenon which possesses processing breakdown

    and backtracking, belongs to the domain of NLU system. GPsentence is a grammatical correct sentence in which originalmisinterpretation lingers until re-decoding has occurred. Anincorrect choice in GP sentence usually is readers' most likelyinterpretation and lures readers into an default parse which,however, finally proves to be a dead end. If the system canpresent the special features of GP sentence, namely,processing breakdown and backtracking, the automaticdecoding can be successful and the effectiveness of systemcan be improved. After showing Octav Popescus model ofNLU system, we emphasize the importance of integration ofsyntactic, semantic and cognitive backgrounds in system,focus on the programming skill of IF-THEN-ELSE statementused in N-S flowchart, and highlight the function of context

    free grammar (CFG) created to decode GP sentence. On thebasis of GP sentences-supported analyses, we come to theconclusions that the machine comprehension can be embeddedin the general frame of communication via programminglanguages, that interaction between the person and machinecan improve NLU system, that programming technology (e.g.N-S flowchart) can help NLU system present the decodingprocedure of GP sentence, and that syntax-supported linguisticskill (e.g. CFG) can bring a deeper understanding of GPsentence.
  • 7/28/2019 IJACSA_Volume3 No11


    (IJACSA) International Journal of Advanced Computer Science and Applications,Vol. 3, No. 11, 2012

    6 | P a g e


    This research is supported in part by grants of YB115-29and YB115-41 from the Eleventh Five-year researchprojects of Chinese Language Application, and grant11YJA740111 from the Ministry of Education and SciencePlanning Project. This research is also sponsored by 2012Yantai Social Science Planning Project Towards theLinguistic Function of the Cultural Development in Yantai: A

    Perspective of Computational Linguistics.


    [1] W. L. Wu, R. Z. Lu, J. Y. Duan, H. Liu, F. Gao, and Y. Q. Chen,Spoken language understanding using weakly supervised learning,Computer Speech & Language, vol. 24, April 2010, pp. 358-382.

    [2] Y. Wilks, A position note on natural language understanding andartificial intelligence, Cognition, vol. 10, July-August 1981, pp. 337-340.

    [3] R. C. Schank, Conceptual dependency: A theory of natural languageunderstanding, Cognitive Psychology, vol. 3, October 1972, pp. 552-631.

    [4] C. Mellish, and J. Z. Pan, Natural language directed inference fromontologies, Artificial Intelligence, vol. 172, June 2008, pp. 1285-1315.

    [5] M. Jeong, and G. G. Lee, Practical use of non-local features forstatistical spoken language understanding., Computer Speech &Language, vol. 22, April 2008, pp. 148-170.

    [6] M. Jeong and G. G. Lee, Machine learning approaches to spokenlanguage understanding for ambient intelligence, Human-CentricInterfaces for Ambient Intelligence, 2010, pp. 185-224.

    [7] K. Hird and K. Kirsner, Objective measurement of fluency in naturallanguage production: A dynamic systems approach, Journal of

    Neurolinguistics, March 2010.

    [8] A. Lockman, and D. Klappholz, The control of inferencing in naturallanguage understanding, Computers & Mathematics with Applications,vol. 9, 1983, pp. 59-70.

    [9] R. Lpez-Czar, Z. Callejas, and D. Griol, Using knowledge ofmisunderstandings to increase the robustness of spoken dialoguesystems, Knowledge-Based Systems, vol. 23, July 2010, pp. 471-485.

    [10] C. Madden, M. Hoen, and P. F. Dominey, A cognitive neuroscienceperspective on embodied language for humanrobot cooperation, Brainand Language, vol. 112, March 2010, pp.180-188.

    [11] K. J. Lee, Y. S. Choi, and J. E. Kim, Building an automated Englishsentence evaluation system for students learning English as a secondlanguage, Computer Speech & Language, May 2010.

    [12] D. Kayser, and F. Nouioua, From the textual description ofan accidentto its causes, Artificial Intelligence, vol. 173, August 2009, pp. 1154-1193.

    [13] J. Bos, Applying automated deduction to natural languageunderstanding, Journal of Applied Logic, vol. 7, March 2009, pp. 100-112.

    [14] W. Minker, Design considerations for knowledge sourcerepresentations of a stochastically-based natural language understandingcomponent, Speech Communication, vol. 28, June 1999, pp. 141-154.

    [15] E. J. Yannakoudakis, I. Tsomokos, and P.J. Hutton, N-grams and theirimplication to natural language understanding, Pattern Recognition,vol. 23, 1990, pp. 509-528.

    [16] O. Popescu, Logic-Based Natural Language Understanding in IntelligentTutoring Systems (PhD Thesis), Carnegie Mellon University, 2005.

    [17] I. Nassi and B. Shneiderman, Flowchart techniques for structuredprogramming, ACM SIGPLAN Notices, vol. 8, 1973pp. 12-26.

    [18] J. L. Du and P F Yu, A computational linguistic approach to naturallanguage processing with applications to garden path sentencesanalysis, International Journal of Advanced Computer Science andApplications. Vol. 3, September 2012, pp. 61-75.

    [19] J. L. Du, P. F. Yu, Syntax-directed machine translation of naturallanguage: Effect of garden path phenomenon on sentence structure,International Conference on Intelligent Systems Design and EngineeringApplications, 2010, pp. 535539.

    [20] A. A. Badawy, Students perceptions of the effectiveness of discussionboards, International Journal of Advanced Computer Science andApplications. Vol. 3, September 2012, pp. 136-144.
  • 7/28/2019 IJACSA_Volume3 No11


    (IJACSA) International Journal of Advanced Computer Science and Applications,Vol. 3, No. 11, 2012

    7 | P a g e

    Extending UML for trajectory data warehouses

    conceptual modellingWided Oueslati#1, Jalel Akaichi*2

    Computer Science Department, Higher Institute of Management


    ABSTRACT The new positioning and information capturetechnologies are able to treat data related to moving objects

    taking place in targeted phenomena. This gave birth to a new

    data source type called trajectory data (TD) which handle

    information related to moving objects. Trajectory Data must be

    integrated in a new data warehouse type called trajectory data

    warehouse (TDW) that is essential to model and to implement in

    order to analyze and understand the nature and the behavior of

    movements of objects in various contexts. However, classical

    conceptual modeling does not incorporate the specificity of

    trajectory data due to the complexity of their components that

    are spatial, temporal and thematic (semantic). For this reason, we

    focus in this paper on presenting the conceptual modeling of thetrajectory data warehouse by defining a new profile using the

    StarUML extensibility mechanism.

    KeywordsTrajectory data; trajectory data warehouse; UMLextension; trajectory UML profile.

    I. INTRODUCTIONThe success of the warehousing process rests on a good

    conceptual modeling schema. In fact, conceptual modelingoffers a higher level of abstraction while describing the datawarehousing project since it stays valid in case oftechnological evolution. Besides, it allows determininganalysis possibilities for the warehouse. However, no

    contribution is at the present time standard in term oftrajectory data semantic models. This finding leads us topropose a new UML profile with user oriented graphicalsupport to represent trajectory data and trajectory datawarehouse conceptual modeling with structural model (classdiagram) and dynamic model (sequence diagram).

    This paper is organized as follows. In section 2, we presentan overview of research works related to conceptualapproaches and extensibility of UML for applications' needs.In section 3, we present the methodology that we adopted toextend the StarUML profile. In section 4, we present theTrajectory UML profile. In section 5, we present the trajectoryUML profile realization. In section 6, we summarize the workand we propose some perspectives that can be done in thefuture.

    II. RELATED WORKSIn this section, we present different approaches related to

    the conceptual modeling methodology, then we presentresearch works that extended UML to adopt it to theirconceptual modeling needs. In the literature, we can find threecategories of conceptual approaches; the top down approach,the bottom up approach and the middle out approach. The

    difference between those latter is situated in the starting point.In fact, each approach has its own starting point such as users'needs, data marts or both users' needs and data marts.Concerning the top down approach, this latter has to answerusers' requirements without any exception. It is very expensivein term of time since it requires the whole conceptualmodeling of the DW as well as its realization and it is difficultbecause it requires the knowledge in advance of dimensionsand facts [1]. In this category, authors of [2] present aMultidimensional Aggregation Cube (MAC) method. Thislatter insures the construction of a multidimensional schema

    from the definition of decision makers' needs but the definedschema is partial because it describes only the hierarchies ofdimensions. The goal of MAC is to supply an intuitivemethodology of data modeling used in the multidimensionalanalysis. It models real world scenarios using concepts whichare very similar to OLAP.

    In MAC, data are described as dimensional levels, drillingrelationships, dimensions, cubes and attributes. Dimensionlevels are a set of dimension members. Those latter are themost detailed modeling concepts and they present real worldinstances' properties. Drilling relationships are used to presenthow one level element can be decomposed of other levels'elements. The dimension paths present a set of drillingrelationships which are used to model a significant sequence

    of drill down operations. Dimensions are used to define asignificant group of dimension paths. This grouping isessential to model semantic relationships. Cubes are the onlyconcept which associates properties' values with realmeasures' values. They insist on the complex hierarchystructure defined by dimensions. The top down approach canbe used in the Goal-driven methodology [3]. In fact, this latterfocuses on the company's strategy by occurring the executivesof the company. For the bottom up approach, this latterconsists on creating the schema step by step (data marts) untilthe obtaining of a real DW [1].

    It is simple to be realized but it requires an important workin the data integration phase. Besides, there is always the risk

    of redundancy due to the fact that each table is createdindependently. Authors in [4] present a dimensional factmodel. This latter relays on the construction of data martsfirstly. This can insure the success in case of complex projectsbut it neglects the role of decision makers. Authors in [5]adopt the bottom up approach. In fact, they present adimensional model development method from traditionalEntity-Relationship models to insure the modelling of DWsand Data Marts. This method is based on three steps: the firststep includes the classification of data models' entities into a

  • 7/28/2019 IJACSA_Volume3 No11


    (IJACSA) International Journal of Advanced Computer Science and Applications,Vol. 3, No. 11, 2012

    8 | P a g e

    set of categories. This leads to the production of a dimensionalmodel from an Entity-Relationship model. We find thetransactional entities that insure the storage of detailsconcerning particular events in the company. We find also thecomponent entities that are directly connected to transactionalentities through the 1..* relationship. Those entities allowdefining details of each transaction. Classification entities areconnected to component entities through the 1..* relationship.Classification entities present the existing hierarchies in themodel.

    The second step consists on identifying hierarchies thatexist in the model. In fact, the hierarchy is an importantconcept in the dimensional modelling level. The third stepconsists on grouping hierarchies and aggregations together toform a dimensional model. At this level, we find two operatorsthat are used to product models of dimensions. In fact, the firstoperator can transform the high level entities to low levelentities. This can be done until the arrival at the bottom of thearchitecture. The aim is to have an only table at the end. Forthe second operator, it is applied on transactional data to createa new entity which contains summarized data. This approachis used as a base for the data driven and user driven

    methodologies. In fact, as presented in [3], the data driven(supply driven) methodology starts by analyzing operationaldata sources to identify existent data. Users' intervention islimited to the choice of necessary data for the decision makingprocess. This methodology is adopted when data sources arevalid.

    For the user driven methodology, it starts by collectingusers' needs. Those needs will be integrated in order to obtainone multidimensional schema. This approach is appreciated byusers but it presents a big challenge. In fact, managers ofprojects must be able to take into account the different pointsof views. For the middle out approach, it is an hybrid methodsince it benefits from the two approaches cited above. Authors

    in [6] present an example of hybrid modelling method that isbased on the top down and the bottom up approaches. Thebottom up approach is based on three steps: the collection ofneeds, the specification and the formalization of those needs inthe form of multidimensional constellation schema. The topdown approach includes the data collection and theconstruction of a multidimensional schema that allowsdecision making. The approach is based on the description ofdecision makers needs. Those two approaches allow havingtwo schemas, then from those latter only one schema will bederived and kept. The middle out approach is composed offour phases; the users' needs analysis, theconfrontation/comparison, the resolution of conflicts and theimplementation.

    Authors in [7] present another method which uses themiddle out approach. This latter is based on three steps: thecollection of users' requirements by the top down approach,the recovery of star schema by the bottom up approach andfinally the integration phase. This latter connects the obtainedstar schema from the first step to the obtained star schemafrom the second step. The integration is realized thanks to aset of matrix. Users' requirements are collected by the GoalQuestion Metric (GQM) paradigm. This latter allowattributing metrics to identified goals. This facilitates the

    filtering and the deletion of not useful goals. Authors of [7]consider that the modeling of warehouses is a process basedon goals, and then users' goals related to DW developmentwill be present explicitly. Goals will be analyzed in order toreduce their number (authors take into account the similarityof goals). For the choice of star schema, authors use theEntity-Relationship model. This latter is exhaustively analyzedto find entities that will be transformed to facts anddimensions. The transformation process of Entity-Relationshipmodel to a star schema is based on three steps. The first step isthe construction of a connected graph that serves tosynthesized data. The second step is to extract a snowflakeschema from the graph. The third step is the integration phase.In fact, authors exploit the structure of the warehouse of thefirst phase and the set of possible schemas of the secondphase, and then they apply a set of steps such as converting ofschema to express them with the same terminology. WithinUML-based conceptual models, the most famous approachesare of Trujillo and his team.

    In [8], authors proposed UML extensions for object-oriented multidimensional modeling. This extension isperformed thanks to stereotype mechanism, tagged values and

    constraints expressed in OCL-Object Constraint Language, inaddition to a set of Well-Formedness rules managing newelements added and determining the semantic of the model.Stereotypes and icons allow an expressive representation ofdifferent constituent elements of a multidimensional modelnamely fact classes, dimension classes, hierarchy levels andattributes. Dimension level classes (stereotyped base classes)should define a directed acyclic graph rooted in dimensionclass. Concerning relationships, the aggregation links facts todimensions, and association/generalization links dimensionlevels (having Base stereotype) between each others.

    In another work of the same team [9], an UML package isproposed to facilitate modeling of large data warehouse

    systems. In fact, they suggest a set of UML diagrams(package) extended with the aforementioned stereotypes, iconsand constraints (OCL) to cope with multidimensionalmodeling and consequently designers will not be limited onlyto the class diagram.

    Several works [10] [11] [12] [13] [14] [15] proceeded byUML profiles to represent their models. In fact, in [11] theproposed profile is used for the oriented agent modeling. In[12] the authors represented a profile for the mobile systemsconception. In [13], authors propose a profile for the modelingof association rules of data mining. In [14] authors propose aprofile to model data mining with the temporal series in thedata warehouse. In [15], authors extended UML to introducenew stereotypes and icons to handle spatial and temporal

    properties at the conceptual level. This led to the visualmodeling tool so called Perceptory.