Knowledge graph representation learning method and system based on text graph enhancement

The invention belongs to the technical field of knowledge graphs, and particularly relates to a knowledge graph representation learning method and system based on text graph enhancement, and the method comprises the steps: carrying out the analysis processing of knowledge graph entity text description, extracting a named entity, and constructing a two-layer heterogeneous text graph formed by sentence layer nodes and text entity layer nodes; establishing a connection between the text graph entity and the knowledge graph entity, obtaining an enhanced knowledge graph, and processing to obtain a node initialization representation; performing semantic propagation among entities by adopting a graph convolutional neural network to obtain entity text representation fusing text content semantics and triple structure semantics; entity text representation is combined with entity structure representation which only considers triples, and updating and optimization are carried out through negative samples and loss function