Intermonth Regularity Classification of Volatile Mobile Internet Users

Penetration of the mobile Internet is increasing worldwide. It amplifies importance of mobile user behavior analysis in research and in industry. The author analyzed a volatile mobile data service based on an assumption that users who visit after a long interval within the same day have a high probability of revisiting the web site in the following month. This is based on a hybrid approach that combines a quantitative measure in mobile click stream mining and behavior modeling of mobile users. Classification of loyal users is done based on the revisit ratio in the following month. A case study of a commercial car information service provides positive results on the classification of loyal users. The author also discusses the classification accuracy from a cross-service comparison perspective.