Learning latent factor from review text and rating for recommendation

In this paper, we propose a model to recommend related products to users. Our model combines the metrits of latent factor model and probabilistic topic model such as latent Dirichlet allocation(LDA), aiming to learn latent user factors from observed reviews rating and latent items factors from reviews text. It provides an interpretable latent factor for users and items. Experiments on a realworld dataset show that our model outperform state-of-the-art methods on the task of recommender system.