基于相似项目与用户评分的协同推荐算法  

Collaborative Filtering Recommendation Algorithm Based on Similar Items and Users Rating

作　　者：陶剑文[1] 潘红艳[1] 

机构地区：[1]浙江工商职业技术学院信息工程系,宁波315012

出　　处：《情报学报》2008年第2期199-204,共6页Journal of the China Society for Scientific and Technical Information

基　　金：浙江省教育厅科研项目资助（20040120）.

摘　　要：推荐系统是电子商务系统中最重要的技术之一。随着电子商务系统用户数目和商品数目日益增加，在整个商品空间上用户评分数据极端稀疏，传统的相似性度量方法均存在各自的弊端，导致推荐系统的推荐质量急剧下降。针对用户评分数据极端稀疏情况下传统相似性度量方法的不足，本文提出了一种基于相似项目与用户评分预测的协同过滤推荐算法，综合利用相似项目和相似用户评分信息预测用户对未评分项目的评分。通过聚类算法形成用户候选近邻集，减小了算法搜索空间，降低了最近邻用户的搜索时间，从而增强了算法的扩展性。实验结果表明，本算法可以有效解决用户评分数据极端稀疏情况下传统协同推荐算法存在的问题，显著提高推荐系统的推荐质量。Recommendation System is one of the most important technologies in E-Commerce. With the development of E- Commerce, the magnitudes of users and Commodities grow rapidly, resulted in the extreme sparsity of user rating data. Traditional similarity measure method works poor in this situation, makes the quality of recommendation system decreased dramatically. To address this issue a novel collaborative filtering algorithm based on similar items and users rating prediction was proposed. This method predicted item ratings that user have not rated by fusing the ratings of similar of items and users. The paper constructed candidated user neighbor sets by using cluster algorithm, which minished the search space and reduced the time of searching the nearest neighbor and consequently boosted up the expansibility of algorithm. The experiment results suggested that this method could efficiently improve the extreme sparsity of user rating data, provided better recommendation results than traditional collaborative filtering algorithms.

关 键 词：推荐系统 协同过滤 相似性 聚类 平均绝对偏差 

分 类 号：F623[经济管理—产业经济] TP391.41[自动化与计算机技术—计算机应用技术]