CF Improvement Based on Probabilistic Analysis of Discrete Explicit Rating Vector

Collaborative Filter (CF) is one of the important algorithms of Recommendation System, the sparsity problem is a significant impediment for real use of CF technique. In this paper, based on probabilistic analysis to users´ discrete explicit rating vector, an All-Average improved algorithm are proposed to solve the problem of CF sparsity and other practical problems. Experimental result show this method improved the precision and quality of CF prediction.