Construction Of Subject Knowledge Graph Based On Deep Learning

It is now an information age.The information on the Internet is gradually used,and the education field is gradually approaching technology and information,making the teaching content more abundant.But for now,the sharing of resources in various disciplines in the field of education is mostly obtained from web pages.The amount of information on the web page is relatively complicated,and it is inevitable that there will be duplication when acquiring resources,and the connection between the knowledge is relatively weak,making the availability of knowledge low.However,the knowledge graph can solve these problems,and it allows learners to more clearly observe the knowledge system of various disciplines and the relationship between various knowledge points in the form of graphs.In this way,knowledge can be better used in teaching.At present,knowledge graphs are widely used in medical,discipline,business and other fields,and play an important role in various fields.The knowledge graph allows us to know the relationship between the entities in the domain more clearly,so it is necessary to construct the knowledge graph,and there are many forms in the construction of the knowledgegraph.In recent years,the method of applying deep learning to the knowledge graph has rapidly developed.This article will use deep learning to construct the knowledge graph of mathematics.In the process of constructing the knowledge graph,the neural network model Bi LSTM+CRF is used to identify entities,and compared with other models.Of Bi LSTM for relationship extraction,and achieved a high accuracy.After entity recognition and relationship extraction,the obtained entities and the relationships between entities are stored in the Neo4 j graph database,and the knowledge graph is drawn.