A natural language recognition method and system

Abstract

The invention relates to a natural language recognition method and a natural language recognition system, wherein the recognition method comprises the following steps that the acquired natural language is divided into a first partition, a second partition and a third partition; according to the first participle, the second participle and the third participle are matched with the pre-established three-dimensional matrix, and the corresponding matching rules are determined and called, so as to realize the recognition of natural language. The three-dimensional matrix comprises a matrix name, a first participle, a second participle, a third participle and matching rules, wherein the first participle is a subject participle, the second participle is an X-axis participle and the third participleis a Y-axis participle, and the X-axis participle and the Y-axis participle intersect to generate various participle combinations. The invention can not only carry out structural treatment on word segmentation so as to save time and labor and facilitate modification, but also use subject word segmentation to set semantic environment, comprehensively sort out related various situations, systematize knowledge points, reduce the error and leakage of word segmentation combination, thereby improving the natural language recognition and processing ability.

Description

technical field [0001] The invention relates to the fields of natural language processing, search engines, intelligent robots, etc., and in particular to a natural language recognition method and system. Background technique [0002] The application of Chinese word segmentation algorithms can be roughly divided into three categories: the first category is based on string matching, that is, scanning a string, and if the substring of the string is found to be the same as the word in the dictionary, it is considered a match, such as the mechanical word segmentation method. This type of word segmentation usually adds some heuristic rules, such as "forward / reverse maximum matching", "long word first", etc.; the second type is based on statistics and machine learning. Features, to model Chinese, that is, to train the model parameters according to the observed data (marked corpus), and then calculate the probability of occurrence of various word segmentations through the model in t...

AI Technical Summary

Problems solved by technology

[0008] In order to overcome the shortcomings of time-consuming and labor-intensive combination of word segmentation, lack of system integration and inconvenient modification of knowledge points in the existing Chinese word segmentation algorithm, the present invention provides a knowledge matrix word segmentation system, which not only performs structural processing on word segmentation, so that Save time and effort, easy to modify, and use topic segmentation to set the semantic environment, comprehensively sort out various related situations, systematize knowledge points, reduce errors and omissions in word segmentation combinations, thereby improving natural language recognition and processing capabilities

Images