A New Collaborative Filtering Algorithm Based on Data Smoothing
Li Ma, Xingjun Wang, Anqi Chen, Riqiang Gao, Yuanyuan Tang, Linghao Xiao · 2017
Based on the advantages of data smoothing, this paper presents a new collaborative filtering algorithm to solve the problem of data sparsity in recommender system. The key innovation of the algorithm consists of clustering and data smoothing. Clustering is used to find out the similarity between users. This paper adopts a new method to cluster users. Data smoothing is designed to solve the problem of data sparsity. It smooths data by introducing Pearson Correlation Coefficient. Then selecting items through nearest neighbor algorithm. Finally, it gets the preference matrix by weighting preference of items. Compared with another typical collaborative filtering algorithm on precision, recall and F1, the new approach performs better.