Implicit discourse relation identification method based on TransS-driven mutual excitation neural network

Abstract

The invention discloses an implicit discourse relation identification method based on a TransS-driven mutual excitation neural network. The method comprises the following steps: (1) constructing an embedding layer of argument and discourse relation; (2) carrying out expression learning of chapter arguments; (3) constructing attention mechanism enhanced representation learning; (4) a sentence translation embedding module (Translating Sentence Embedding, TransS); (5) constructing a chapter relationship identification module; and (6) constructing a mutual excitation mechanism. According to the method, firstly, an argument pair-relation embedding layer is used for obtaining an embedding vector of an argument pair and a relation, then distributed representation of the argument pair and the relation is modeled through an argument pair-relation encoding layer and introduction of an attention mechanism, and finally representation parameters are optimized and the relation recognition performance is improved through mutual guidance between TransS and a relation recognition module.

Description

technical field [0001] The invention relates to the technical field of discourse analysis in natural language processing, in particular to the discourse relationship recognition technology, in particular to an implicit discourse relationship recognition method based on a TransS-driven mutual excitation neural network. Background technique [0002] A discourse relation describes how two adjacent text units (e.g., clauses, sentences, and larger sentence groups) are logically connected to each other, and is usually defined as a conjunction with two arguments (Arg1 and Arg2, respectively) . Explicit textual relations can be easily identified with an accuracy of about 94%, while implicit textual relations are identified without explicit connectives. Therefore, implicit discourse relation identification remains a challenging problem, which requires relation inference from specific contexts. Implicit discourse relationship recognition is beneficial to many popular Natural Languag...

AI Technical Summary

Problems solved by technology

[0007] The entity translation embedding model (TransE) is an effective method to predict the missing relationship between entities in the knowledge graph, which models the relationship by interpreting the entity relationship as the translation operation of the entity in a low-dimensional vector space [9], that is, if ( h e , l e ,t e ) is established, it is consistent with the tail entity vector t e should be close to the head entity vector h e plus the relation vector l e , but has not been effectively applied at the sentence level

Images