Probabilistic Latent Factor Model for Collaborative Filtering with Bayesian Inference
This paper provides an incremental improvement for collaborative filtering recommendation systems by addressing overfitting in Latent Factor Models.
This paper addresses the overfitting issue in Latent Factor Models (LFM) for collaborative filtering, which arises from sparse user-item interactions. The authors propose Bayesian Latent Factor Model (BLFM) and its extension BLFMBias, which introduce regularization through prior constraints on latent factors and use Variational Inference for prediction. Experiments on a movie rating dataset demonstrate the effectiveness of their models.
Latent Factor Model (LFM) is one of the most successful methods for Collaborative filtering (CF) in the recommendation system, in which both users and items are projected into a joint latent factor space. Base on matrix factorization applied usually in pattern recognition, LFM models user-item interactions as inner products of factor vectors of user and item in that space and can be efficiently solved by least square methods with optimal estimation. However, such optimal estimation methods are prone to overfitting due to the extreme sparsity of user-item interactions. In this paper, we propose a Bayesian treatment for LFM, named Bayesian Latent Factor Model (BLFM). Based on observed user-item interactions, we build a probabilistic factor model in which the regularization is introduced via placing prior constraint on latent factors, and the likelihood function is established over observations and parameters. Then we draw samples of latent factors from the posterior distribution with Variational Inference (VI) to predict expected value. We further make an extension to BLFM, called BLFMBias, incorporating user-dependent and item-dependent biases into the model for enhancing performance. Extensive experiments on the movie rating dataset show the effectiveness of our proposed models by compared with several strong baselines.