Optimization collaborative filtering recommendation algorithm based on ratings consistent
Wei Ze, Dengwen Zhou · 2016
In view of the traditional based on user collaborative filtering algorithm has the neighbor set selection is not accurate.1So propose a new similarity measure method, named score based on consistent with the collaborative filtering recommendation algorithm. The algorithm selects the user common rating data to calculate the user's similarity, considering the common score the consistency of the data in the user item rating, constructed evaluation matrix, and rating consistent times than scoring item number as a penalty function is introduced into the similarity calculation, which bring the nearest neighbors of user, and predict the item's rating to recommend, alleviate the similarity calculation value and actual value deviation. Experimental results show that, the improved algorithm proposed in this paper significantly increase the prediction accuracy, so as to improve the quality of recommendation.