Community Based User Behavior Analysis on Daily Mobile Internet Usage

Laptops, handhelds and smart phones are becoming ubiquitous providing (almost) continuous Internet access and ever-increasing demand and load on supporting networks. Daily mobile user behavior analysis can facilitate personalized Web interactive systems and Internet services in the mobile environment. Though some research have already been done, there are still some problems need to be investigated. In this paper, we study the community based user behavior analysis on the daily Mobile Internet usage. What we focus on in this paper is to propose a framework which can calculate the proper number of the clusters in mobile user network. Given a mobile user Internet access dataset of one week which contains thousand of users, we firstly calculate the hourly traffic variation for the whole week. Then, we propose to use cluster coefficient and network community profile to confirm the presence of communities in mobile user network. Principal Component Analysis (PCA) is employed to capture the dominant behavioral patterns and uncover the several communities in the network. At last, we use communities/clusters to work out the various interests of the users on the timeline of the day.