Document-level entity relationship extraction method based on document structure and external knowledge

Abstract

The invention discloses a document-level entity relationship extraction method based on a document structure and external knowledge. The method comprises the steps: obtaining a document text, constructing a structure diagram corresponding to the document text according to the document text, initializing the nodes and edges of the structure diagram, and obtaining a structure diagram initialization result; and based on a structure chart, the structure chart initialization result and the trained edge-oriented graph neural network model, obtaining an updated edge, and inputting the updated edge into a classifier to obtain an entity relationship extraction result of the document text. According to the embodiment of the invention, the entity relationship extraction result is obtained through the method, the performance of document-level relationship extraction is improved, the problem of imbalance of positive and negative examples is solved, and the method has better capability of processing partial knowledge missing and has better expandability.

Description

technical field [0001] The invention relates to the technical field of artificial intelligence, in particular to a document-level entity relationship extraction method based on document structure and external knowledge. Background technique [0002] With the advent of the era of information explosion, information extraction plays an important role in the processing of massive unstructured text data. Relational extraction is an important part of information extraction. It is widely used in knowledge graphs, information retrieval, question answering systems, sentiment analysis, and text mining. It aims to extract two entity pairs <entity 1, entity 2> Specific types of information, so as to output structured triple information <entity 1, relationship, entity 2>, which is a bridge between entity extraction and event extraction. [0003] Existing research on relation extraction is mainly aimed at relation extraction at the sentence level. Traditional relation extract...

AI Technical Summary

Problems solved by technology

[0008] The technical problem to be solved by the present invention is to provide a document-level entity relationship extraction method based on document structure and external knowledge, aiming at solving the need to manually construct corresponding features in the document extraction method in the prior art. , which is time-consuming and labor-intensive, and does not have scalability. The method of adjusting hyperparameters is time-consuming and labor-intensive. It is not convenient to deal with the shortcomings of lack of knowledge and low scalability, and cannot fully utilize the role of external knowledge.

Images