Multi-modal knowledge map is adopted,YOLOv3-tiny model is used for target detection training,and the medical knowledge graph is constructed by top-down and bottom-up methods.The intelligent teaching platform uses the MVC three-tier model,MySQL and Neo4j databases,and Elasticsearch is used as the search engine database.The experiment shows that the fifth training of the research model significantly improves efficiency.In medical teaching image search,the model recogni-tion accuracy reaches 81.0％,the running time is 2.7 seconds,and the loss function drops to about 0.25.The practice proves that the multi-modal knowledge graph constructed is very successful,and the intelligent teaching platform is used to realize the efficient search of medical teaching images,which provides the basis for the intelligent development of medical education.
关键词
医学教育/多模态知识图谱/智能教学
Key words
medical education/multi-modal knowledge graph/intelligent teaching
维普
万方数据