20 / 2017-12-22 14:12:29

A Collaborative Filtering Recommendation Based on Clustering and Common Liking Rate

Sparse Data; common liking rate; sparse subspace; collaborative filtering

In order to reduce the negative impacts of sparse data, a collaborative filtering recommendation method based on sparse subspace clustering and common liking rate is proposed. Firstly, the sparse subspace clustering method is used to cluster the users, and initial filling of the user's rate data, so that more useful information can be retained. Then, select the user's common scoring data, two users with little difference in rates are selected as the common liking rate set.The similarity of users is calculated according to the common liking rate set. The common liking rate set can better reflect the user similarity and reduce the error.At last, search the nearest neighbor users and generate recommendation result set.The experimental results on real data sets show that the algorithm can predict the user's rate more effectively in the case of sparse data.