基于用户分类的隐含因子模型研究*

针对现有隐含因子模型存在的新用户和项目的冷启动问题,提出基于用户分类的隐含因子模型,将用户分类信息融入到隐含因子的矩阵分解当中,先在原评分矩阵和用户分类信息的基础上使用指示函数和数据归一化等方法构建一个分类评分矩阵,再将分类评分矩阵融入到隐含因子模型的评分预测中。通过与传统隐含因子模型等方法在多个不同隐含因子个数上的实验比较分析,实验结果表明,改进模型能够不仅能解决新用户和项目的冷启动问题,还能有效降低预测评分的均方根误差,并提高预测推荐的准确度。

Research on latent factor model based on user classification

To solve the problem of user cold start and item cold start, a kind of latent factor recommendation algorithm based on the user classification was proposed in which user classification information is integrated into matrix factorization. A classification rating matrix was created firstly by using the method of index function and unifying based on the user-item rating matrix and user classification information, and then the data of classification rating matrix was combined into the rating prediction stage in the latent factor model. A number of latent factors were used to compare the new method with the traditional latent factor model. The experimental results showed that the proposed model could not only solve the cold start problem, but also could reduce the root mean squared error between predicted ratings and real ones, and improve the accuracy of prediction effectively.