基于项目聚类的协同过滤推荐算法

Collaborative Filtering Recommendation Algorithm Based on Item Clustering

推荐系统是电子商务中最重要的技术之一 ,协同过滤是推荐系统中采用最为广泛也是最成功的推荐技术 .随着电子商务系统用户数目和商品数目日益增加 ,在整个用户空间上寻找目标用户的最近邻居非常耗时 ,导致推荐系统的实时性要求难以保证 .针对上述问题 ,本文提出了一种基于项目聚类的协同过滤推荐算法 ,根据用户对项目评分的相似性对项目进行聚类 ,生成相应的聚类中心 ,在此基础上计算目标项目与聚类中心的相似性 ,从而只需要在与目标项目最相似的若干个聚类中就能寻找到目标项目的大部分最近邻居并产生推荐列表 .实验结果表明 。

Recommendation system is one of the most important techniques used in E Commerce. Many recommendation systems employ collaborative filtering to generate recommendations. With the gradual increase of users and commodities in E Commerce, the time consuming nearest neighbor search of the target user in the total user space resulted in the failure of ensuring the real time requirement of recommendation system. A collaborative filtering recommendation algorithm based on item clustering was proposed in this paper to solve this problem. Items were clustered based on users ratings on items, each cluster has a cluster center. Based on the similarity between target item and cluster centers, the nearest neighbors of target item can be found in the item clusters that most similar to the target item. Experimental results indicated that this algorithm could effectively improve the real time performance of recommendation systems.