基于Web客户因子分析的协同推荐算法

Collaborative recommendation algorithm based on Web customer factor analysis

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,提出了基于客户因子分析的协同推荐算法。该算法利用因子分析将客户向量进行降维处理,得到几个具有代表性的客户因子,然后用这些客户因子对目标客户进行回归分析,进而预测目标客户对待评项目的评分值。最后通过实验证明了该算法的有效性,为以后研究推荐算法提供了一种新的途径。

In order to solve the problem that data overload and data sparsity in collaborative filtering recommendation algorithm,this paper proposed a collaborative recommendation algorithm based on customer factor analysis.The algorithm reduced the dimensions of customer vector by using of factor analysis and got some representative customer factors,which used to regression analysis of target customer to forecast the evaluated items.Finally,experiments show that the algorithm is effective,which provides a new way for future recommendation algorithm research.