Rating prediction using review texts with underlying sentiments

Recommender systems typically produce a list of recommendations to precisely predict the user's preference for the items. For this purpose, latent factor models, such as matrix factorization, are usually employed to find latent factors that can characterize both users and items by observed rating scores. Recently, online user feedback accompanied with review texts has become increasingly common. The review texts contain not only users' attention to the situation of the different aspects of items, but also users' sentiment towards different aspects of specific items. However, traditional latent factor models often ignore such review texts, and therefore fail to characterize users and items precisely. Furthermore, although some current studies do employ review texts, many of them do not consider how sentiments in reviews influence the rating scores. In this paper, we propose an extended Hidden Factors as Topics Model (HFT) (a model combining the Latent Factor model and the Latent Dirichlet Allocation) based on Aspect and Sentiments Unification Model (ASUM) (an extended topic model), called Ratings Are Sentiments (RAS). By combining users' sentiments in review texts and their rating scores, our model can learn more precise latent factors of users and items compared with the baseline models. The extensive experiments on large, real-world datasets demonstrate that the RAS model performs better than both the latent factor model and the HFT model and alleviates the cold-start problem to some extent.