Utilizing User Feedback Data for Content Recommendations

Abstract

Nowadays, Internet contains massive amount of information. In this environment, people who seek specific information could be overwhelmed by the options that they can reach through the Internet. To help users filter the information and overcome the information overload problem, recommender systems play an important role. Here, we deal with a specific recommendation problem – recommending content to users in a content management system utilizing users’ feedback data. We have tried both content-based and collaborative filtering approaches. In the content-based approach, once the content profile is built, user profile could be built based on different categories of user feedback data. We have explored the effect of these different feedback categories on the recommendation result. In the collaborative filtering approach, the feedback data is used for building the user-content rating matrix and matrix factorization is then applied. The experiment result shows that content-based approach outperforms collaborative filtering approach for this particular problem.