A text relationship extraction method and system based on a hierarchical knowledge graph attention model

Abstract

The invention provides a text relation extraction method and system based on a hierarchical knowledge graph attention model, and the method comprises the steps: firstly selecting a training text set and a knowledge graph according to a to-be-processed text, and carrying out the preprocessing of the training text set and the knowledge graph; Then constructing a hierarchical knowledge map attentionmodel, and training the model by using the preprocessed training text; And finally, marking the head entities and the tail entities of all sentences in the to-be-processed text, and inputting the marked to-be-processed text into the trained hierarchical knowledge map attention model to obtain a relation result. The method has the advantages that the hierarchical knowledge graph attention model isprovided according to the technical scheme, the knowledge graph is used for representing the weights of sentences and words distributed to the model, the accuracy rate and the recall rate of relationprediction of the model are improved, and then the extraction accuracy of text relations is improved.

Description

technical field [0001] The present invention relates to the field of relation extraction, in particular to a text relation extraction method and system based on a layered knowledge map attention model. Background technique [0002] Knowledge bases provide effective structured information for real-world facts and are used as key resources by many natural language processing (NLP) tasks such as web search and knowledge question answering. Typical knowledge graphs include Freebase, DBpedia and YAGO. Since the facts in the real world are arguably endless and growing every day, existing knowledge graphs are far from complete. Therefore, the task of information extraction is valued by more and more people. [0003] Information extraction aims to extract structured information from large-scale unstructured or semi-structured natural language texts. Relation extraction is one of the important sub-tasks, the main purpose is to extract the semantic relationship between entity pairs...

AI Technical Summary

Problems solved by technology

[0007] Although the current relationship extraction method using remote supervision has achieved good results, the problem of wrong labels generated during the alignment process has always been troubled.

[0008] Traditional models of remote supervision rely heavily on experts in specific knowledge fields to manually design features, or use natural language processing (NLP) annotations such as part-of-speech tagging and syntactic analysis to provide classification features. Obviously, manual design of features is too time-consuming and laborious, and NLP tools often exist Many errors, such as named entity recognition (NER), dependency analysis, etc., the more feature engineering will bring more errors, the propagation and accumulation of errors will occur in the pipeline of the entire task, and ultimately affect the accuracy of subsequent relationship extraction

Images