The Design And Implement Of Recommender System Based On Knowledge Graph Embedding

Due to the explosive growth of information,recommender systems have been playing an increasingly important role in online services.The relevant research of it has a high practical value.Collaborative filtering(CF)based methods,which make use of historical interactions or preferences,have made significant success.However,CF methods usually suffer from limited performance when user-item interactions are very sparse.To tackle these problems,this thesis proposes a recommender model based on text and knowledge graph.Considering the strong ability of feature extraction of deep learning in text processing and efficient knowledge encoding of represent learning,this thesis proposes a collaborative filtering recommender model based on text summary using deep learning and knowledge graph embedding.This thesis mainly focuses on using the knowledge graph and deep learning in the recommender model.The specific research content is as follows:(1)This thesis utilize text summary based on deep learning algorithm to encoding text information and proposes a collaborative filtering recommender model based on text summary using deep learning.The data sparsity is relieved by utilizing the text information in collaborative filtering(2)Based on(1),this thesis utilize a new knowledge graph embedding and proposes a collaborative filerting recommender model based on text summary and knowledge graph embedding.(3)Experimental results on a CiteUlike dataset verify that our approach outperforms several state-of-the-art methods with significant advances.(4)Based on the above work,an online thesis recommender system was designed and implemented.