中文摘要：

针对协同过滤问题，提出了一种基于高斯混合的概率模型，称为真实偏好高斯混合模型．用户对项目的评分由三个因素决定：用户对项目主题和内容的真实偏好，用户的评分习惯，以及项目的公众评价．引入了两个隐含变量，分别用于描述用户类和项目类，用户和项目依概率可以同时属于多个类．模型包括离线建模过程和在线预测过程，在线预测可以在常数时间内完成．实验表明新模型的预测结果明显优于其他几种协同过滤算法．

英文摘要：

This paper presents a new probabilistic Gaussian mixture model for collaborative filtering, named real preference Gaussian mixture model （ RPGMM ）. In RPGMM, the user＇ s vote to an item is determined by the user＇s real preference to the item, the rating style of the user, and the public praise of the item. There are two latent variables corresponding to classes of user and item. Each user or item may be probabilistically clustered to more than one groups. The parameters of the model can be estimated offline, and the online predictions can be computed in constant time. Empirical study has shown that our new model outperforms several other collaborative filtering models and algorithms remarkably.