协同过滤算法中推荐集选取方法的研究

Study on the selection method of recommendation in Collaborative Filtering Algorithm

针对传统的基于余弦相似性的协同过滤算法中推荐集选取方法进行了改进,设计了一种新的评分方式预测用户对未评价项目的评分,从而增强了推荐的合理性。实验结果表明,该算法同传统协同过滤算法相比能显著提高推荐精度。

In this paper, the traditional cosine similarity based on the collaborative filtering algorithm to select sets recommend methods to improve the design of a new prediction of the score the way users evaluate the project did not score, thus enhancing the rationality recommend.Experimental results show that the algorithm with the traditional collaborative filtering algorithms can significantly improve the accuracy of recommend.