Offering Program Recommendations through Social Network Services

摘要

The increasing popularity of social network services (SNS) has prompted many users to post daily events and happenings on SNS. More and more people share their thoughts on programs while watching TV. The NHK Science & Technology Research Laboratories has proposed a program recommendation technique based on postings about programs and relationships among SNS users. The goal is to encourage people to watch programs by recommending them scheduled broadcast programs that are likely to be watched by their friends and acquaintances or popular SNS users (hereafter called "charismatic users"). That is, the idea is to exploit the user tendency of "It sounds good since so-and-so with similar tastes is going to watch it." An SNS enables users to form relationships. For example, Twitter users can register favorite users to "follow" (Figure 1). For users who use the favorite function, our system can recommend programs similar to those favorite users previously viewed by doing an analysis of user-posted information from their favorite users (Figure 2(a)).