Collaborative Filtering in Recommendation Systems: Idea and Evaluation

추천 시스템의 협업 필터링: 아이디어와 평가

Collaborative filtering has been used frequently as a recommendation system. To reduce the errors on predicting the ratings that may be given by the user, we propose a new aggregation method to do so. We used a real-world dataset MovieLens to compare our proposed method from previously existing methods, and accordingly to the results, ours was more accurate.