Research on Recognition Method of Textual Implication

Abstract

Text implication recognition aims to judge the semantic and logical relationship between two paragraphs of text. The reasoning process involves syntactic analysis, vocabulary understanding, logical reasoning, social experience, and common sense. It is a way to judge whether the computer is to a certain extent. The challenging research task of "understanding" text semantics is also one of the more important benchmark tasks in the field of natural language processing. It is found that in the past ten years, the total number of publications in this field has been increasing year by year, and the popularity is also increasing. At present, there are mainly text implication relationship recognition methods based on similarity, text implication relationship recognition methods based on alignment, and deep neural networks. The textual implication relationship recognition method. This article summarizes the different methods.