The Research on Collaborative Filtering Recommendation Algorithm Based on Improved Clustering Processing

In applications of personalized recommendation, user similarity of common clustering algorithms only considers user relationship without considering relationship between users and items, the similarity above reduces the accuracy of clustering, making it difficult to find similar users, and the same with item similarity. This paper improves the distance function of data clustering algorithm by Hamming distance, making accuracy of clustering much higher, so running Slope one on the processed data set above improves accuracy of recommendation significantly.