| Summary: | 碩士 === 國立臺北科技大學 === 經營管理系碩士班 === 105 === The rise of Internet has cultivated a new type of television viewing behavior. Many viewers watch TV programs and simultaneously participate in discussions on social media such as Facebook, and therefore, it has become significant for the employees in TV industry to learn how to improve ratings in terms of Facebook. In this research, we take an example of Baseball World Cup and use Text Mining to identify keywords that affected TV program ratings from the posts and the comments in Facebook. Then, CART and VAR are applied to build up a TV program rating forecasting model based on those identified keywords. According to the result, the built model is capable of making good prediction. It, to a certain extent, captures the ratings tendency and provides TV industry an approach to understand what kind of keywords are really helpful for their program ratings.
|
|---|
