Adaptive user profile from mobile phone based positionaing using data mining technique | |
| Author | Nattawut Yangkhruea |
| Call Number | AIT Thesis no.RS-18-08 |
| Subject(s) | Location-based services--Social aspects Mobile communication systems Data mining |
| Note | A thesis submitted in partial fulfillment of the requirements for the degree of Master of Engineering in Remote Sensing and Geographic Information System, School of Engineering and Technology |
| Publisher | Asian Institute of Technology |
| Abstract | In the Location Based Service Applications it is important to know user's profile in order to provide service to the right person. However, user's profile should build as fast as possible. Trajectory data from a mobile phone device can discover user's profile such as, significant place, transportation mode and daily mobility pattern. First, significant places extraction need to identify related features and calculation then related all features was imported and stored in PostgreSQL database server and was accessed the model by using Psycopg2 module from Python as an adapter. Then Random forest algorithm with hyper-parameter tuning was employed to classify significant place such Home, Work and Other. Secondly, transportation mode detection was analyzed by calculated related features including total point, total distance, total time, minimum velocity, maximum velocity, maximum acceleration, average and velocity. Then Random Forest was applied to classify transportation mode including Car, Motorbike and Train. Third, detecting daily mobility pattern of user and identify the shortest period to detect the main mobility pattern. Splitting data was applied, then generate origins and destination (Ol'i) and calculate the percentage of each Ol). Then visualize Ol) in Gephi was applied to finding the main pattern. Finally calculation the average shortest time and the percentage of main pattern was utilized. Finally, the significant places extraction, transportation mode detection and daily mobility pattern with the minimum time was applied to the test data. The result shown that the minimum time of extracting significant place such as Home is 5 days, Work Place and Other is 7 days. For transportation mode detection, the minimum time is 1 day since it can classify when related features ready. Moreover, the average minimum time of building main daily mobility pattern is 11 days |
| Year | 2018 |
| Type | Thesis |
| School | School of Engineering and Technology |
| Department | Department of Information and Communications Technologies (DICT) |
| Academic Program/FoS | Remote Sensing (RS) |
| Chairperson(s) | Apichon Witayangkurn |
| Examination Committee(s) | Sarawut Ninsawat ;Miyazaki, Hiroyuki |
| Scholarship Donor(s) | Royal Thai Government |
| Degree | Thesis (M. Sc.) - Asian Institute of Technology, 2018 |