Coinonia: Privacy-Aware Meeting Scheduler based on Location … · 2014. 5. 27. · the meeting scheduler. In Section 4, we present how to implement the application that meets design

Coinonia: Privacy-Aware Meeting Scheduler based onLocation-Based Services for Mobile Users

Hyoungtae [email protected]

AbstractThe location based service has been attracted moreand more in recent years with the advance of wire-less communication technologies. Some mobile appli-cations are designed to provide users with their cur-rent location information shared by a group of peoplewhere a user might be reluctant to reveal the privateinformation. The paper presents a meeting schedul-ing application, Coinonia to address a way of con-trolling activities in time to deal with the conflict be-tween benefits of utilizing location information andrisks of violating privacy in location tracking tech-nologies. This privacy-aware application integratesGoogle Map views with the features of meeting sched-uler, which enables meeting members to gather to-gether with less effort by having them communicateefficiently.

Keywords: location-based service, mobile applica-tion, privacy, meeting scheduler

1. Introduction

A various set of new technologies in mobile devicesand social networks has brought the rapid developmentof mobile applications. The new generation mobilephones called smartphones allow users more advanc-ing computing powers to support the common featuresof computer or PDA as well as traditional phone-callfunctionalities. Among these features, location-basedservices have become more prevalent since a hand-held device is one of tools used for tracking locationand it prompts users for location information.

These mobile applications greatly enhance our qual-ity of life with regards to providing users with neededinformation in the right place and at the right time.Such location systems, for examples, offer location-based services to give a driving direction to the near-est restaurants or gas stations for nearby residents ortourists. Location-based social networking servicesare also available to get in touch with not only friends

and family but also interesting people of common in-terests. Moreover, they enable college students to no-tify campus policemen when they are in trouble. De-spite the benefits of location-based technical supports,individual privacy is threatened by revealing the user’scurrent location implying what he/she is doing to un-wanted people or in undesirable time.

In this paper, we present a mobile application,Coinonia to share user location information with meet-ing members to efficiently get together. The appli-cation organizes meeting schedules inserted by users,shows the current locations of people who participatein the meeting during a certain time – monitoring time,and makes them communicate one another by sendinginstant messages. Once the meeting begins, the loca-tion services of this application is automatically ter-minated with a consent of meeting creator to protectlocation privacy. It also provides users with alarmspersonally set up in a mobile phone: 1) late alarm bycomputing the estimated arrival time to the meetingplace and 2) message alarm on receiving a messagefrom others.

The remainder of the paper is organized as follows.The next section briefly reviews the previous worksrelated to our mobile application. In Section 3, we dis-cuss motivations and background concepts to designthe meeting scheduler. In Section 4, we present how toimplement the application that meets design conceptsand requirements. Finally, Section 5 summarizes ourcontributions and proposes future works.

2. Related Work

As a number of commercial mobile application havelaunched, the features of web-based scheduling pro-grams such as Google Calendar[3] or HopStop[5] areintegrated into mobile applications. Above all, how-ever, the benefit of using these mobile applications isto utilize the user location information based on thestatic data inserted by users. Developments in mo-

1

bile geographical positioning and location-based ser-vices have made it easier for mobile users to lookup a physical location of interesting points or peo-ple nearby. For example, TimeToGo[2] is an applica-tion that combines personal calendars with public tran-sit options and reminds users about upcoming eventswith departure alerts. This smart reminder systemaggregates the components of diary, public transporttimetable, and automatic alerts. Another location-based service[8] over the phone recommends popularsocial events around a user. Besides the personal useof location-based services, sharing location informa-tion like Google Latitude[4] or Mologogo[7] becomespopular as well.

Despite the benefits of mobile applications withlocation-based services, some researches[6, 11] havebeen conducted to examine privacy concerns in track-ing location information. Among such applications toconsider the privacy issue, Connecto[1] is an locationsharing application that is concerned with privacy byallowing users to manually enter the name of locationshared by a small social group to hide their actual loca-tion. Another example, Locaccino[10] provides usersrich privacy controls set by group, time, and locationsto resolve privacy issues. It also shows an interest-ing analsys that high entrophy locations, namely fre-quently visited places are less private by users.

Although these location-aware applications and theirresearches help to enhance location sharing privacywith more sophisticated settings, they have a limita-tion in using for a meeting scheduler to share locationinformation with formal groups in most cases, not so-cial groups including family members or close friends.While most of other mobile location services focuson an individual user’s privacy preferences, Coinoniatakes care of how the value of location information canbe maximized in an agreeable manner among meetingmembers.

3. Application Design

We develop a mobile meeting scheduler, Coinonia tomake our life easier, which provides users with loca-tion services of the meeting place and members as wellas scheduling services in the perspective of informa-tion dynamics.

3.1. Motivation

Meeting is an important part of our daily lives sincewe continuously meet someone to communicate each

other, to share information, or to make a decision. Wewrite down a meeting schedule not to forget when,where, and with whom to meet. Moreover, until everymeeting member arrives at a meeting, we contact oneanother to make sure if he/she will arrive on time orto leave a message to other members. Let us considera scenario that one person wants to check the currentlocations of other members. It requires one-to-manyconversations.

Figure 1. One person needs one-to-many conversa-tions to check other members’ locations

On the other hand, an mobile application with locationservices over the smartphones makes it much easier,replacing one-to-many conversations for each personwith one Map representation showing user geograph-ical locations. Clearly, sharing location informationamong meeting members may help them to gather to-gether.

Nonetheless, if a user’s privacy cannot be preservedin using the location service after a meeting ends sothat his/her current location is monitored by a groupof people other than family members or close friends,the user will hesitate to use the application. Assign-ing a set of different privacy settings to meeting mem-bers would not be preferred since it reflects personalfavor to affect relationships with group members. Re-garding that logging off from the system is not easyto remember whenever the meeting is over, the mon-itoring functionality of this application is guaranteedto be automatically terminated when it is no longerneeded.

3.2. Information-centric approach

In order to design a meeting scheduler, we take aninformation-centric view that explicitly consider therole of information and its dynamics in our system.Agrawala[9] claims that every piece of information hasvalue within a given context related to its use and/or

2

purpose and the value of information depends on time.In addition to his claim, the value of information mayincrease when it is given at right place and in right timefor providing location information since a user’s cur-rent location will also dynamically change in location-based services as time passes. Therefore, our appli-cation is designed to have a functionality of notifyingwhether or not a user will be late to a meeting basedon estimated arrival time against meeting time.

Furthermore, his statements give us an insight to re-solve the privacy issue in our information system. Forinstance, users can have more benefits sharing locationinformation than costs of giving other their private lo-cation information to other meeting members within acertain time before the meeting. Once a meeting be-gins which means no location information is needed,the costs of sharing information become greater. As aresult, the system should terminate location services atthe end instead of remaining connected.

Why does the system have to take care of the termi-nation instead of leaving it to a user’s privilege? It’sbecause the idle users tend to forget their log-off fromthe service as we can see the cases in school computersand because some reserved users would be anxious tobecome the first person to stop sharing personal infor-mation with others. Coinonia asks a meeting creatorto terminate a service of meeting scheduler when themeeting begins, and makes the service completed onehour after the scheduled meeting time assuming that itwould be rarely used anymore for a meeting purposeunless the meeting needs to be rescheduled.

3.3. System Architecture

In Figure 2, We explain system flows of Coinonia intwo cases: a)when a meeting schedule is inserted bya meeting creator and b)when meeting members use itto meet together.

When a meeting creator creates a meeting scheduleand Coinonia, its meeting information is inserted intothe server through the HTTP server. In the mean-while, the notification of creating the meeting is sentto each meeting member. As soon as member con-firms the message, the application updates informationabout members with their confirmation, and starts tomonitor a user’s location for local late alarms based onhis/her alarm setting. Within a monitoring time, theaggregated location information of monitored personsis shared by the meeting group, and they are allowed

Figure 2. System Layout

to send and/or receive messages through the applica-tion.

4. Implementation

We develop and test our application on an emulatorin Android 2.2 platform implemented in Java. It isan open source and enables Google Map features. Tolook up and store the data, we use a MySQL databaseserver and communicate with it via the Apache HTTPserver since our application cannot directly connect toMySQL through Android. Coinonia provides regis-tered ID users with two main functions as a meetingscheduler: (1)Scheduling process and (2)Map Rep-resentation of user locations. They are accompaniedwith the instant messaging and meeting alarm.

4.1. Login Page

The first page of the application is Login page thatrequires a correct combination of UserID and Pass-word. This process allows only registered users to ac-cess to the information system and to manipulate meet-ing schedules.

4.2. Main Page

When passing through the Login page, the Main Pagein Figure 3 appears with two tabs – Schedules with alist of meetings that a user participates in and Mes-sages with a list of received messages. Initially, alist of meetings scheduled on today’s date comes out.Each row in the list consists of meeting title, time, and

3

Figure 3. Login page and Meeting List in Main page

a green/yellow/red/black circle indicating a possibil-ity of arriving on time such as safe/hurry/late/alreadypast respectively. In our example, a meeting scheduleLate Night Study at 9:30pm appears with a green cir-cle since there was enough time left for the meetingcomparing to its current time 8:35pm, while time foranother meeting schedule Dinner has already passedso that Dinner has a black circle. When it becomes9:26pm which is 4 minutes before the meeting, thecolor of circle was yellow since the user was nearenough to arrive the meeting place (in about 11 min-utes).

The details of color coding is shown as follows:

color time rangegreen timer > timea +max(0.2 ∗ timea, 10)red timer < timea −max(0.2 ∗ timea, 10)

yellow between green and redblack timer < 0

Table 1. Color coding for late status

In this table, timer is a remaining time for the meet-ing time and timea is an estimated arrival time fromthe current user location to the meeting place usingGoogle APIs. All numbers are in minutes and roundedup. At present, computing arrival time is based onmoving by car but it could be improved with moretransit options.

The text message at the top frame includes the currentuser’s ID and the current selected date for the meet-ing schedules. Among three buttons below, the usercan navigate between dates using Set Date button andlook through a list of meetings of the date. The restof buttons are for creating a new meeting schedule and

for alarm setting, and we will discuss them in the nextsections(Section 4.3 and 4.6).

4.3. Create Page

The Add button in the Main page brings a user to theCreate page where he or she can generate a new sched-ule. In this process the person becomes a creator of themeeting. He describes its details, decides which (bothID and without ID) members are participated, and setsup monitoring time/privacy settings. Once a meetingis created, each ID member in the list receives a notifi-cation and confirms to share his/her location informa-tion. Let us consider creating Late Night Study as anexample.

Figure 4. Create Page

The meeting date and time is initially set to the currentones so that the application prevents the user from di-rectly inserting date and time in a wrong format. Mod-ification can be done only by clicking on EDIT buttonfollowed by date/time changes in a dialog box.

As for the meeting place, Coinonia allows users to setthe location with pointing it out on the Google Map.Before setting it up, they are able to check its ac-tual address by pressing on A(address) button. Zoomin/out functionalities and Satellite views are also avail-able and all the corresponding function keys may befound in Help menu. Next, Coinonia obtains mem-ber information from users. Once Members sectionare filled out, the system recognizes which membersare ID(registered) members or not and Coinonia willmonitor locations of only ID members with confirma-tion in sharing their location information with other

4

members.

In the privacy setting, each user is allowed to set uphis/her privacy setting with regards to location andtime. Whether or not to share Location and Monitoringtime will be initially determined by a meeting creator.Location privacy setting has two levels: (1)Estimatedarrival time from the current location to the meetingplace and (2)the current Location where the user is lo-cated in. By default, Estimated arrival time must bechecked out. How long a meeting member will bemonitored by other members before the meeting be-gins is also selected on creating a meeting. When de-tails of the new meeting are filled out, it will be in-serted into the meeting list in Main page by CREATEbutton.

4.4. Details Page

If a user selects one of meeting list items in Main page,Coinonia enables a user to see its specific informationsuch as date, time and place in the Details page. Fig-ure 5 shows an example when the first meeting sched-ule Late Night Study is selected. The Estimated Ar-rival time is calculated for each member based on bothhis/her current location and the meeting place.

As we mentioned earlier, there are two groups of mem-bers – ID members with confirmation in using this lo-cation service and any other members – and the maprepresentation with Map button represents only the lo-cation information of the first group marking with thecolor-coded pushpins. In this example, we can no-tice that a meeting member away from the meetingplace(marked with a blue icon) was indicated by a redpushpin from a green one as the meeting time getscloser.

Figure 5. Detail Page and Map representations

If a pushpin is clicked, whom it represents and theuser’s brief message to other members, if any, areshown up. This map representation with location in-formation saves time and effort to grasp the othermembers’ location and situation on the way to themeeting place.

4.5. Message Communication

If users move onto Message tab at right top cornerin Main page, they will see a list of received mes-sages from others. The application provides a messen-ger function to efficiently communicate each other. Amember can broadcast the current late status or leavea message to group members so that it may avoid one-to-many conversations as with map representation. Ofcourse, one-to-one messages such as writing a mes-sage to a particular ID and replying it are possible. Aswith a meeting list, when one of message items is se-lected, its details come out (Figure 6) and a user canreply the message back to the sender.

Figure 6. Message list and its Details page

For a particular meeting, if we click on Messages but-ton in meeting Details page in Figure 5, we couldwrite down a message to one of meeting member orall the members. The instant message function helpsCoinonia’s efficient communication besides map rep-resentations.

4.6. Alarm Setting

Through the Alarm button in Main page, the alarm set-ting can be made personally. It allows users to haveflexibility in receiving information for late status andnotice of other members’ messages. While the privacysetting is applied on the revelation of information tothe entire meeting group, the alarm setting is locally

5

applied on notification of his/her own late status andwhether or not to receive messages.

5. Conclusion

In this paper, we present a mobile application,Coinonia to provide location based services preserv-ing privacy. It is equipped with meeting schedulerfunctionalities and Google Map representation so thata group of meeting members easily gather together at ameeting place in terms of saving communication effortto find each other as well as notifying the user’s currentlocation. Time is a key variable to resolve a privacy is-sue by automatically terminating location informationsharing among members as meeting starts.

At present, the further development of our mobile pri-vacy preserving meeting scheduler focuses on two as-pects: (1)testing Coinonia over smartphones in realenvironment settings to enhance its functionalities and(2)considering more transit options.

References

[1] BARKHUUS, L., BROWN, B., BELL, M.,SHERWOOD, S., HALL, M., AND CHALMERS,M. From awareness to repartee: sharing loca-tion within social groups. In Proceeding of thetwenty-sixth annual SIGCHI conference on Hu-man factors in computing systems (New York,NY, USA, 2008), CHI ’08, ACM, pp. 497–506.

[2] CAMERON, R., AND GRANTHAM, A. Time-togo: A mobile application for departure alertsbased on google calendar, transit and gps data.In UbiComp (2009), ACM.

[3] GOOGLE. About google calendar. http://www.google.com/googlecalendar/new.html.

[4] GOOGLE. Google latitude mobile. http://www.google.com/mobile/latitude/.

[5] HOPSTOP. Hopstop announces more mo-bile options! http://www.hopstop.com/?action=newsletter&path=http://www.hopstop.com/img/news20101216.html.

[6] MICHAEL, K., MCNAMEE, A., AND

MICHAEL, M. The emerging ethics of hu-mancentric gps tracking and monitoring. InProceedings of the International Conference on

Mobile Business (Washington, DC, USA, 2006),IEEE Computer Society, pp. 34–.

[7] MOLOGOGO. Mologogo. http://www.mologogo.com/.

[8] QUERCIA, D., LATHIA, N., CALABRESE, F.,LORENZO, G. D., AND CROWCROFT, J. Rec-ommending Social Events from Mobile PhoneLocation Data. In Proceedings of the 10th

IEEE International Conference on Data Mining(ICDM) (Sidney, Australia, December 2010).

[9] RONALD, A. A., LARSEN, R. L., AND SZA-JDA, D. Information dynamics: An information-centric approach to system design. In Proceed-ings of the International Conference on VirtualWorlds and Simulation (2000).

[10] TOCH, E., CRANSHAW, J., DRIELSMA, P. H.,TSAI, J. Y., KELLEY, P. G., SPRINGFIELD,J., CRANOR, L., HONG, J., AND SADEH, N.Empirical models of privacy in location sharing.In Proceedings of the 12th ACM internationalconference on Ubiquitous computing (New York,NY, USA, 2010), Ubicomp ’10, ACM, pp. 129–138.

[11] WANG, J. L., AND LOUI, M. C. Privacy andethical issues in location-based tracking systems.In Proceedings of the 2009 IEEE InternationalSymposium on Technology and Society (Wash-ington, DC, USA, 2009), ISTAS ’09, IEEE Com-puter Society, pp. 1–4.

6