Inferring substitutable and complementary products with Knowledge-Aware Path Reasoning based on dynamic policy network

Inferring the substitutable and complementary products for a given product is an essential fundamental concern recommender system. To achieve this, existing approaches take advantage of knowledge graphs to learn more evidences inference, whereas they often suffer from invalid reasoning lack elegant decision making strategies. Therefore, we propose novel Knowledge-Aware Path Reasoning (KAPR) model which leverages dynamic policy network make explicit over graphs, inferring relationships. Our contributions can be highlighted as three aspects. Firstly, this inference scenario Markov Decision Process in order accomplish knowledge-aware path graphs. Secondly, integrate both structured unstructured provide adequate accurate decision-making. Thirdly, evaluate our on series real-world datasets, achieving competitive performance compared with state-of-the-art approaches. code released https://gitee.com/yangzijing_flower/kapr/tree/master.