The K-Nearest neighbor slope one algorithm based on weighted user similarity and user tag

Slope One algorithm has been widely used in collaborative filtering algorithm. However this algorithm does not consider about the similar relationship among users, and the accuracy of the algorithm is affected. Combining with the idea of the K-Nearest Neighbor, this article weights fusion Improved Jaccard and Pearson ways to bring in the user neighbors to calculate the deviation terms in Slope One algorithm, and the prediction accuracy is improved. And then the prediction accuracy is further increased by bringing the user tag information. In the end, applying the algorithm to the MovieLens Data Sets, the simulation results show that the prediction accuracy of the algorithm in this paper is better than the traditional Slope One algorithm and some other algorithms.