一种协同过滤方法及其在信息推荐系统中的实现  

A Collaborative Information Filtering Method and Its Application in Information Recommending System

作　　者：姜丽红 [1] 徐博艺 [2] 张海燕 [3] 

机构地区：[1]上海交通大学计算机系,上海,200030 [2]上海交通大学管理学院,上海,200233 [3]宁夏大学数学计算机学院,银川,750021

出　　处：《情报学报》2005年第6期669-673,共5页Journal of the China Society for Scientific and Technical Information

基　　金：中国科学院资助项目

摘　　要：本文提出了一种基于模糊聚类技术的协同过滤方法,应用模糊聚类技术从项目的属性特征上对项目进行聚类,用隶属度的值来表示项目属于每个模糊簇的程度,由用户-项评分矩阵和模糊簇的隶属度值,构造用户-模糊簇的偏好矩阵,进而利用用户-模糊簇偏好矩阵获得用户相似群体,为用户实现基于协同过滤的文档推荐.最后,利用过滤技术实现了一个科技文献推荐系统,对多种推荐策略进行了验证.In this paper the research of information filtering is firstly reviewed. A fuzzy-cluster-based collaboration filtering algorithm is proposed. In the algorithm the fuzzy-clustering method is applied to cluster the items characters. The grade of membership is used to describe which group the item belongs to. The User-Cluster profile matrix is composed of the User-ltem profile matrix and the grade of membership. The documents recommending are made based on the User-Cluster profile matrix. Finally one science document recommending system is designed using filtering methods. In the system multi-filtering methods are used and tested.

关 键 词：协同信息过滤 模糊聚类 信息推荐系统 隶属度 

分 类 号：G201[文化科学—传播学] TP311.5[自动化与计算机技术—计算机软件与理论]