考虑项目依赖不对称性的协同过滤推荐方法

Collaborative filtering recommendation method considering asymmetry of items' dependence level

为了降低用户-项目评分矩阵稀疏性对实验结果的影响,提出了一种基于项目依赖度不对称性的协同过滤算法。在分析用户-评分矩阵的基础之上,定义了一种基于项目支持度的计算公式,结合传统的相似度计算公式,设计了基于项目依赖度的计算公式,对比实验结果表明,新方法提高了推荐的准确度,降低了稀疏性对实验结果的影响,取得了较好的效果。

To reduce the influence of recommendation result caused by the User-Item rating matrix sparse, a cooperative filter algorithm based on the asymmetric of item dependency is put forward. Firstly, on the basis of analyzing the User-Item rating ma- trix, a project based support for the asymmetry between the calculation formula is designed, and combining with traditional simi- larity formula, the item dependency formula is defined. Finally, the comparative experimental results show that the proposed method achieves a better precision of the recommendation.