Neural Collaborative Filtering for Chinese Movies Based on Aspect-Aware Implicit Interactions

Aspect information mining from user comments has become an important means to improve the performance of recommendation systems (RSs). This is because aspect in fine-grained and tends reflect interactions preferences users over items multiple dimensions. These are different ratings, which often explicit linear. Most current RSs based on learn contribution aspects a linear manner, while ignoring implicit features non-linear aspects. Since Chinese grammar greatly with English grammar, there few models movie comment In this work, we propose architecture, named aspect-based neural collaborative filtering (ANCF), extract terms rules formulated dependency parsing. The proposed ANCF integrates generalized tensor factorization tensorized multi-layer perceptrons into network capture user-item-aspect mixed nonlinear way. potential interaction vector actual layered fused processing, can reduce sparsity solve cold start problem certain extent. Performance results show that model outperforms some traditional ones accuracy effectiveness.