A Collaborative Filtering Recommendation Algorithm Improved by Trustworthiness
Shengjun Xie · International Journal of Future Generation Communication and Networking · 2014
Recommender systems based on collaborative filtering have been well studied in both industry and academia fields.However, traditional collaborative filtering methods are typically in a low accuracy and lack of resistance towards attacks such as non-reliable information.To this end, in this paper, we propose a collaborative filtering recommendation algorithm improved by trustworthiness.Specifically, first we employ a content based method to identify a set of users with similar interests.Then, a trust model is applied as a scoring function, and higher ranked neighbors are selected as the evidence of prediction.Besides, we conducted extensive experiments using Netflix dataset, and the results show that our method is more efficient compared with others.