SNS trend-based TV program recommendation scheme

Recently, as countries have switched over from analog broadcasting to digital broadcasting, various services for digital broadcasting have been introduced to enhance customer satisfaction. However, users still experience difficulties in selecting what they want to watch due to the large volume of TV contents. Often, social trends and TV programs are closely related; TV programs could be produced to deal with some hot issues and public interest could be formed for some TV program. Nowadays, such public interest can be detected as trends from popular SNS such as Twitter. In this paper, we propose a SNS trend-based TV program recommendation scheme. To do that, we first extract trend keywords from Twitter stream data and augment them semantically by referring to recommendation log and portal sites. Finally, we analyze EPGs to find matched TV programs to the trends. We implemented a prototype system to evaluate the effectiveness of our scheme.