PEV: A new item similarity measure for Item-Based collaborative filtering

In Item-Based collaborative filtering recommendation algorithm, the most critical component is the similarity calculation between items. By analyzing the weakness of traditional similarity measures in data sparsity, this paper presents a new item similarity measure. for Item-Based collaborative filtering. The new measure takes synthetically into account the influence of user rating on item similarity from three factors of-Proximity, Effect and Value. The experiment shows that the new measure can effectively avoid the shortcomings of traditional similarity measures under the data sparsity condition, reduce the negative effect on the final recommendation, and can provide better recommender results for the system.