基于双向主题模型的协同过滤算法

Dual Collaborative Topic Regression for Recommendation Systems

主题模型可以学习用户和推荐项目的潜在主题分布。提出了一种基于双向主题模型的协同过滤算法,分别学习用户和推荐项目的潜在主题分布用于推荐服务。在真实的数据集上实验验证,该算法的性能均优于几个经典的协同过滤算法。

Topic model can be used to learn the latent topic distribution. A new collaborative filtering al- gorithm based on dual collaborative topic regression to learn the user＇s latent topic distribution and the item＇s latent topic distribution for recommendation is proposed. On a large real-world dataset, the experi- ment results illustrate that the approach achieves a better performance than the state-of-the art collabora- tive filtering methods.