On Analyzing Geotagged Tweets for Location-Based Patterns

Abstract

Geotagged social media is becoming highly popular as social media access is now made very easy through a wide range of mobile apps which automatically detect and augment social media posts with geo-locations. In this paper, we analyze two kinds of location-based patterns. The first is the association between location attributes and the locations of user tweets. The second is location association pattern which comprises a pair of locations that are co-visited by users. We demonstrate that through tracking the Twitter data of Singapore-based users, we are able to reveal association between users tweeting from school locations and the school type as well as the competitiveness of schools. We also discover location association patterns which involve schools and shopping malls. With these location-based patterns offering interesting insights about the visit behaviors of school and shopping mall users, we further develop an online visual application called Urbanatics to explore the location association patterns making use of both chord diagram and map visualization.

Publication
Proceedings of the 17th International Conference on Distributed Computing and Networking - ICDCN ‘16