Recommendation system knowledge graph link prediction method and device, equipment and medium

• Main Authors: ZHANG XUANYU, WU JIANBIN
• Format: Patent
• Language: Chinese; English
• Published: 12.04.2024
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; DATA PROCESSING SYSTEMS OR METHODS, SPECIALLY ADAPTED FORADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORYOR FORECASTING PURPOSES; ELECTRIC DIGITAL DATA PROCESSING; PHYSICS; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE,COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTINGPURPOSES, NOT OTHERWISE PROVIDED FOR

The invention discloses a recommendation system knowledge graph link prediction method and device, equipment and a medium. Comprising the following steps: counting the importance of each entity in a positive example triple of the knowledge graph, and randomly replacing the entities with low importance to obtain a negative example triple; constructing a knowledge graph data set based on the positive example triad and the negative example triad of the graph; constructing a knowledge graph link prediction model comprising a mapping hyperbolic space module, a graph convolution module and a link prediction module; and inputting the knowledge graph data set into the knowledge graph link prediction model for training, and inputting the knowledge graph to be predicted into the trained knowledge graph link prediction model for link prediction. According to the method, the negative example is obtained based on the entity relationship quantity and random number comparison strategy, and the embedded vector information is