A TV Series Recommender Svstem Based On Implicit Feedback

The age we are living now is an age of information explosion. The data on Internet is always increasing rapidly with the development of network techniques and online applications. A consensus among people is that we should use recommender to solve this kind of data overload. As for the field of TV series, it also can’t keep out of the wave brought by Internet.A new kind of TV series information website is becoming more and more popular now. Compared with traditional TV series websites, these kind of websites don’t have any copyright and can’t provide user activity log on how each user watch TV series. However, these websites focus on a special TV series categories and provide professional service. As a result, this kind of websites can be a good sample in the research of recommender system.In this paper we propose a TV series recommender system based on implicit feedback. More specifically speaking, our recommender can not only deal with explicit feedback such as whether user is interested in a TV series, but also implicit feedback such as how user check information about each TV series. At the same time, this recommender can balance data in a proper way. Finally, the precision of our recommender system can be improved in these two ways.We designed several experiments to test our recommender system and the result shows it can work well.