Entity relationship extraction method and equipment

Abstract

The invention provides an entity relationship extraction method and equipment. The method comprises the following steps: acquiring an entity pair data set containing a preset relationship, wherein theentity pair data set comprises a plurality of entity pairs; extracting sentences containing the entity pairs from professional data in the medical field; screening out a sentence template used for representing the relationship from the sentences based on an initial BERT model; and adjusting the initial BERT model based on the sentence template and the entity pair data set so as to perform relationship extraction on the entity pair data set of which the relationship is to be extracted through the adjusted BERT model. Through the scheme, the data is extracted from the professional data in the medical field, and the sentence template for representing the relationship is screened out based on the extracted sentences, so that the working efficiency of data annotation and feature matching is improved, the method can adapt to various data in the medical field, a large amount of manpower is not needed, and the cost is saved.

Description

technical field [0001] The present invention relates to the technical field of data relation extraction, in particular to a method and device for extracting entity relation. Background technique [0002] At present, for medical entities such as diseases and symptoms, diseases and operations, it is necessary to extract the relationship between entities, and there are two existing extraction methods: one is based on rule extraction, and the other is based on supervised learning. The method based on rule extraction is to use pre-established rules to extract corresponding entities from the text or to judge whether the entities conform to the corresponding relationship, while the method of supervised learning is to train the classifier to judge after marking a large amount of data. Whether there is a corresponding relationship between entities. Both of the current solutions have some problems: [0003] The method of using rules depends on the quality of the rules, and requires ...

AI Technical Summary

Problems solved by technology

[0003] The method of using rules depends on the quality of the rules, which requires a lot of manpower input in the early stage, and the rules made manually may not cover all types of relationships, and the recall rate is poor

The method of using supervised learning requires a large amount of labeled data, which is costly, time-consuming and labor-intensive, and if a new relationship needs to be extracted, the data needs to be labeled again, which is less flexible.

Images