A Regularity-Oriented Service Model with Sub-Day-Scale User Classifications for Mobile Internet

The Internet has been rapidly mobilized by increasing penetration to wireless users. Important issues in mobile user behaviors include identification of long-term regular users in dynamic 24-hour behaviors. The author proposes a regularity-oriented service model to outline the mobile- specific aspects of usage regularity to obtain clues to user preferences on content and services. Then, the author performs some user surveys to identify the diversity of mobile Internet use. It brings our attention on user segment awareness to identification of regularity. The author proposes a clustering method based on an assumption of three different segments of mobile users. The author examines the clustering result with the revisit ratio in the next month to demonstrate the usefulness of the method in the context of long- term use regularity.