基于评分支持度的最近邻协同过滤推荐算法

针对传统协同过滤推荐算法存在推荐质量不高的局限性,提出一种基于评分支持度的最近邻协同过滤推荐算法。该算法用调整后的共同评分次数动态调节相似度的值,以更真实地反映彼此间的相似性。然后计算目标用户和目标项目的最近邻集合及各自评分和支持度,根据评分支持度自适应调节基于目标用户和目标项目的评分对最终推荐结果影响的权重。与其他算法的对比实验结果表明,该算法能有效避免传统相似度度量方法存在的问题,从而提高了推荐质量。

Collaborative filtering recommendation algorithm based onnearest-neighborhood and rating support

To solve the shortcomings of the traditional collaborative filtering recommendation algorithms, this paper proposed an improved collaborative filtering recommendation algorithm for the nearest neighbors based on rating support. First on the basis of correlation similarity, this algorithm adopted an improved similarity measure method which could dynamically adjust the value of similarity according to the modified common rating. Then, computed predicting rating and rating support of the active user and item based on the nearest neighbor sets. Finally, according to the rating support data, adjusted different self adaptive influence weights of the neighbor sets of the active user and the active item, and obtained the final recommendation results. The experimental results show that compared with the other recommendation algorithms, the algorithm can effectively avoid the defects of traditional similarity measure and improve the recommendation quality.