Exploring urban characteristics using movement history of mass mobile microbloggers

Abstract

The explosive growth of the smartphones market and social network sites such as the Twitter has enabled to share numerous short messages reflecting our daily lives and social events from outside their homes or offices. Compared to conventional blogging sites, micro-blogging sites through mobile network enable us to easily write and share their daily logs without any spatial or temporal restrictions. Such mass geo-tagged and time-stamped micro-blogs can inform us about social patterns, regardless of their scale, time, or significance. We investigate characteristic patterns in urban areas such as populated areas from the movement histories of mass mobile micro-bloggers. In particular, some interesting movement patterns can be frequently observed in urban areas using our two measures such as aggregation and dispersion. We also present experimental results in determining urban characteristics from actual micro-blog dataset from the Twitter.