Movie Recommendation System Based on Semantic Web and Machine Learning

It is difficult to acquire, from these massive fragmentary data, valuable design knowledge for designers in conceptual product design. The paper introduces a new approach to design knowledge acquisition through the integration of deep learning with knowledge graph techniques. In particular, through the knowledge extraction model, design-related entities and their relations can be extracted from fragmentary data, and then the knowledge graph is constructed to support design knowledge acquisition for conceptual product design. Apart from that, the knowledge extraction model proposes ALBERT to resolve the problems of memory limitation and communication overhead in the entity extraction module, and applies multi-granularity information to segment errors and polysemy ambiguity in the relation extraction module. Experimental comparison verified that the proposed knowledge extraction model was effective and accurate. In this case study, it could prove that the knowledge graph construction is feasible with real fragmentary porcelain data. More than this, it has shown capability in providing designers with connected and visualized design knowledge.