The ever-changing nature of the mobile Internet contributes to the difficulties encountered in user behavior research. Regularity is an important aspect of the mobile Internet in research and marketing, because it end users easily lose their interest and leave the mobile web sites due to the limited visibility of the web. Maintaining user loyalty is a vital challenge for mobile webs. Therefore, the methods to identify loyalty transitions are important. The author proposes a regularity measure using click counting in the time slots. With the assumption that the users with regular access have more chances to continue to use the mobile webs, the author examines the monthly prediction of user behavior based on the user access regularity in the previous month. The author obtains approximate 80 % accuracy of prediction in the case study. The author discusses the limitation and implications of the comparison.
展开▼
