Implementation and Performance Evaluation of a Mobile Music Recommender

There have been active studies on the processing of ordinal scale-based ratings in CF due to shortcomings of cardinal scale-based ratings. These studies assume an explicit rating scheme for capturing customer preferences. In the mobile Web, however, customers find it hard to directly rate his/her taste on some products because of the mobile device’s poor interfaces. Therefore, an implicit rating method to represent customer’s preference in the ordinal scale is necessary. In this paper, we propose a mobile music recommendation system that implicitly creates ordinal scale-based customer profiles using mobile Web mining technique. An experiment with the participation of real mobile Web customers shows that the proposed system provides better performance than existing CF systems in the mobile Web environment.