本文已被：浏览 1937次 下载 3708次
Received:November 08, 2010 Revised:December 05, 2010
Received:November 08, 2010 Revised:December 05, 2010
中文摘要: 为解决协同过滤推荐算法中的数据量过大和数据稀疏性的问题，提出了基于项目因子分析的协同推荐算法。该算法通过采用因子分析将项目向量降维为几个具有代表性的项目因子，然后用这些项目因子对目标项目进行回归分析，进而预测目标客户对待评项目的评分。最后通过实验证明了算法的有效性，为以后研究推荐算法提供了一种新的途径。
Abstract:In order to solve the problem that data overload and data sparsity in collaborative filtering recommendation algorithm, the collaborative recommendation algorithm based on item factor analysis is proposed in this paper. The algorithm reduces the dimensions of item vector by use of factor analysis and gets some representative item factors, which are used to regression analysis of target items to forecast the evaluated items. Finally, experiments show that the algorithm is effective, which provides a new way for future recommendation algorithm research.
文章编号： 中图分类号： 文献标志码：
基金项目:辽宁省教育厅科学技术研究项目(W2010212);教育部博士点基金项目(200801470004);教育部人文社科基金(10YJC630407)
引用文本：
赵宏霞,王新海,杨皎平.基于项目因子分析的Web 客户需求协同推荐算法.计算机系统应用,2011,20(7):188-191,210
ZHAO Hong-Xia,WANG Xin-Hai,YANG Jiao-Ping.Collaborative Recommendation Algorithm of Web Customer Demand Based on Item Factor Analysis.COMPUTER SYSTEMS APPLICATIONS,2011,20(7):188-191,210
赵宏霞,王新海,杨皎平.基于项目因子分析的Web 客户需求协同推荐算法.计算机系统应用,2011,20(7):188-191,210
ZHAO Hong-Xia,WANG Xin-Hai,YANG Jiao-Ping.Collaborative Recommendation Algorithm of Web Customer Demand Based on Item Factor Analysis.COMPUTER SYSTEMS APPLICATIONS,2011,20(7):188-191,210
