Entity relation joint extraction method and system based on relation guidance

Abstract

The invention belongs to the technical field of natural language processing, and particularly relates to an entity relation joint extraction method and system based on relation guidance, sentences in a target text are encoded, and sentence vector representation in the target text is obtained; for the sentence vector representation, a relation extraction module is used to extract a relation type contained in the target text; the extracted relation type is used as priori knowledge to be fused with word vector representation in a target text sentence, and an entity recognition module is used for recognizing an entity corresponding to the extracted relation type in the target text. Attention to irrelevant entities can be reduced, extraction of redundant entities is avoided, then entity pairs corresponding to multiple identified relation types are identified, the problem of entity overlapping is solved, all entity relation triples contained in sentences are extracted finally, entity relation identification accuracy is improved, and actual scene application is facilitated.

Description

technical field [0001] The invention belongs to the technical field of natural language processing, and in particular relates to a method and system for joint extraction of entity-relationships based on relationship orientation. Background technique [0002] With the advent of the era of big data, a large amount of data is generated on the Internet all the time, and most of these data exist in unstructured form. How to transform massive unstructured data into structured information is currently a concern. . Information extraction technology came into being under this background. Information extraction technology refers to the technology that automatically extracts information such as events, entities, and relationships from natural language texts, and outputs them in a structured form. Entity relationship extraction is a subtask of information extraction, the purpose is to identify entities in the text, and the relationship between entities, and represent them in the form ...

AI Technical Summary

Problems solved by technology

Although entity recognition and relationship classification can choose models independently and freely, which has strong flexibility, this type of method has some disadvantages: error propagation, in the case of inaccurate entity recognition, will affect the next relationship classification task; Ignore the intrinsic connection between the two tasks of entity recognition and relationship classification; generate redundant entity pairs, because the entities of the entities need to be paired, and then classify their relationship types, N entities will generate N2 entity pairs , a large number of entity pairs are irrelevant, and these redundant entity pairs will affect the relationship classifier

Images