This research topic can only be taken by Thesis route and Project route student. This research topic include the following project development. The project outcome needs to be examined by well-designed experiment for quanititative and qualitative data analysis (for Thesis route student) and pilot testing for usability and technology acceptance analysis (for both Thesis and Project students).
User wandering behaviours may involve many location visits in different order. The research team has proposed an algorithm which can provide users recommendation for their next visit according to the similarity of their behaviours between each others and the connections amongst locations. In order to test the effectiveness of proposed algorithm the research team also has developed a mobile app – Next Stop Recommender – for Android platform. The mobile app and the web user interface can collect sequential positioning data of travelers. The server side Java servlets can find the visit lists of travelers and summarize popular route patterns; moreover, it can turn the route patterns into recommendation rules and make recommendations for travelers.
The research has three limitations which are the objectives of this research project. The first limitation of this research is that the proposed algorithm currently treats all visited places equivalent without taking the categories of the places into consideration. In such case, a non-relevant place like gasoline station or grocery store visited in the middle of a camping trip may keep the system from finding more accurate popular patterns, finding recommendation rules and making appropriate recommendations. The second limitation of this research is that the important places and its area are manually entered into the system now. The system should be capable of importing the point of interests directly from open access map data – Open Street Map. With such enhancement, the travelers in world wide will benefit from the research. The third limitation of this research is the lack of considering social effects while making recommendations for travelers. For instance, a traveler may change his or her mind to pay a particular place a visit because he or she finds hundred thousands of other travelers went to there before or hundreds of travelers 'Like' the place. As the system is now capable of being aware of whether the recommendations are taken by the travelers or not, it can let the traveler know how many travelers before him or her had accepted the particular recommendation. In such case, the social effects of a recommendation can be measured and the recommendation algorithm can be revised to a more adaptive one to provide travelers better personalized service.
Tasks of this project also include:
The student can learn how to design an Android app. Also, the student needs to consider how to make the app still useful when the internet access is not available; therefore, appropriate synchronizers and multi-agent system may need to design and implement as well. New relevant functions have to be developed in modularized JSPs to give flexibility and reusability in future development and enhancement for working with other systems. A pilot study is needed to verify the effectiveness of the proposed algorithm and the usability of the Next Stop Recommender app.
In this project, the student needs to design and develop the following functions in corresponding web and Android platform:
1. replace the use of google map with Open Street Data so the Next Stop Recommender can be used when users have no Internet access
2. map data pre-fetch function on Android platform
3. refine the database that Next Stop Recommender currently has to make the point of interests (and its categories), map objects, map tags, and map data available for the system
4. daemon of updating (download and upload map data, objects, and tags) Open Street Data for Next Stop Recommender server
5. map annotation function for users to add and to see map tags and objects as well as the covered area on the Web platform
6. design and implement algorithm to identify the travel behavior such as walking, running, driving (take bus and subway OR driving), etc. that a user may have so the system can give more appropriate recommendation based on the findings
7. design and implement function to do data cleaning and enrichment so the system can have better performance in making recommendations and the users can see more smooth route on the web UI
8. design and implement algorithm to identify the visit and stay behavior that a user may have while his or her device can not receive GPS signal
9. refine the algorithm of making recommendations by taking point of interests and its categories into consideration
Learning outcomes:
The student can learn how to design a multi-agent based Android app as well as distributed services/agents in modularized JSPs. Also, the student can know how to design a pilot (or a complete experiment for thesis students) to gather the perception that users may have toward the system and to do usability analysis for the system (in addition, for thesis students, they can learn how to design research model, make hypotheses, verify the causal relationships amongst factors, do quantitative and qualitative data analysis). Skills of designing and evaluating a questionnaire's validity and reliability can be learnt either.
For taking this topic, the student is better to have the following skills or knowledge:
1. Familiar with XML, JSON, HTML, CSS, Java Script, and AJAX
2. Having experience in developing Android applications
3. Having experience in developing application in Java standalone application (acts as startup daemons running in Debian OS), JSP, and Java servlets
4. Information collection, re-organization, summarization, and reports
5. Using screen capture software and video editing software/tool to creating instructional video clips |
