A Methodology of Personalized Recommendation System on Mobile Device for Digital Television Viewers

With the increasing of the number of digital television (TV) channels in Thailand, this becomes a problem of information overload for TV viewers. There are mass numbers of TV programs to watch but the information about these programs is poor. Therefore, this work presents a personalized recommendation system on mobile device to recommend a TV program that matches viewer’s interests and/or needs. The main mechanism of the system is content-based similarity analysis (CBSA). Initially, the viewer defines favorite programs, and then the system utilize this list as query to find their annotations on the WWW. These annotations will be used to find other programs that are similar by using CBSA. Finally, all similar programs are grouped to the same class and stored as a dataset in a personal mobile device. For the usage, if a TV program matches the interest and specified time of viewer, the system on mobile device will notify the viewer individually.