Matrix factorization for recommendation with explicit and implicit feedback

Matrix factorization (MF) methods have proven as efficient and scalable approaches for collaborative filtering problems. Numerous existing MF methods rely heavily on explicit feedback. Typically, these data types may be extremely sparse; therefore, these methods can perform poorly. In order to address these challenges, we propose a latent factor model based on probabilistic MF, by incorporating implicit feedback as complementary information. Specifically, the explicit and implicit feedback matrices are decomposed into a shared subspace simultaneously. Then, the latent factor vectors are jointly optimized using a gradient descent algorithm. The experimental results using the MovieLens datasets demonstrate that the proposed algorithm outperforms the baselines.