How do you predict your future movement and locations?

• Main Author: Ko, Zhan Hong
• Other Authors: School of Computer Engineering
• Format: Final Year Project
• Language: English
• Published: 2014
• Subjects: DRNTU::Engineering::Computer science and engineering
• Online Access: http://hdl.handle.net/10356/59033
• Institution: Nanyang Technological University

This project serves to provide an analysis on GPS data and external information like the calls and messages of a person to predict his/her future movement and location. To do so, an Android application was created to collect location, call and message information on the mobile user. The data collection process spanned over a total duration of two and a half months from the 1st of January to the 15th of March. On a daily basis, the Android application named “STracker” has obtained user locations from 00:30 hours to 23:30 hours. The raw data collected was then processed on a remote server that has been set up to generate the frequented locations visited by the user. It also serves to reduce the load on the mobile system for concerns of resource usage and data integrity. From the processed data, the routine of the user has also been obtained after performing analysis on the various temporal aspects and these routine locations include areas of Jurong, Choa Chu Kang, Tampines, and Bukit Batok. The “Home” and “School/Work” areas have also been identified. A list of the user’s visits to abnormal locations were then extracted by removing the routine locations and further analysis is performed on the data. It has been found that there exists patterns that relates the user’s message and call frequencies to the probability of making a visit to an abnormal location. The range of distance that the user usually travels was then used to predict the likely range of areas that the user would be on the subsequent future abnormal locations.