Dependency analysis model and Chinese joint event extraction method based on dependency analysis

Abstract

The invention discloses a Chinese joint event extraction method based on dependency analysis. The method comprises the following steps: firstly, introducing dependency analysis to construct a syntactic structure to enhance deep interaction of information; secondly, three types of edge representations are designed for bridging the inconsistency of words so as to calculate graph convolution features; finally, through joint learning of an event trigger word classification task and an event argument classification task, the problem of cascade error propagation of a traditional pipeline type method is relieved, and the effect of extracting event trigger words and arguments from documents is improved. According to the Chinese joint event extraction model based on dependency analysis, syntactic structure information is fused while semantics are coded, information flow between words is enhanced, and different types of edge representations are designed for constructing an undirected graph according to the characteristics of Chinese word segmentation. According to the method, semantic feature representation is enriched by fusing syntactic structure knowledge contained in a Chinese text, and the sentence-level event extraction effect is effectively improved by using a joint learning method.

Description

technical field [0001] The invention belongs to the research field of event extraction of information extraction in natural language processing, and particularly relates to a dependency analysis model and a Chinese joint event extraction method based on dependency analysis. Background technique [0002] The rapid development of the Internet has brought a huge amount of data, the information contained in it has great value, and how to quickly locate the required information in hundreds of millions of data has always been an urgent problem to be solved. There are various carriers of data, such as sound, image, text and so on. As the most important way of human communication, the study of language helps machines to understand human intelligence, and thus the field of natural language processing was born. Text and language are inseparable. Most of human knowledge is recorded and handed down in the form of language. Therefore, the study of text has always been the core part of a...

AI Technical Summary

Problems solved by technology

[0005] The technical problem to be solved by the present invention is to provide a dependency analysis model and a Chinese joint event extraction method based on dependency analysis, which solves the problem that the event argument extraction algorithm involving syntactic structure information in the prior art involves less influence on data processing accuracy and efficiency. The problem

Images