A method and system for extracting semantic information similar to natural language based on combination theory

Abstract

The invention discloses a method and system for extracting natural-language-like semantic information on the basis of the combinatorial theory. The system comprises a rule developing module, a rule combining module and a data matching module. The rule developing module is used for developing a rule system according to an existing body, an existing semantic dictionary and an existing classifying system and semantic rules are defined in the writing mode similar to that of the natural language. The rule combining module is connected with the rule developing module and used for combining matching training sets according to the rule system to generate a mechanism allowing semantic rules to be selectable. The data matching module is connected with the rule combining module and used for conducting data matching on a target text according to the matching training sets and obtaining semantic information of the target text. The method and system for extracting the natural-language-like semantic information based on the combinatorial theory have the advantages that the semantic information is abundant, the accuracy rate and the recall rate are high, the cost is low and industrialization can be realized.

Description

Technical field [0001] The present invention relates to the technical field of information extraction, in particular to a natural language-like semantic information extraction method and system based on combination theory. Background technique [0002] Information extraction is to extract specific real-time information from text. In most information extraction, the semantic information of the text is determined by identifying entities in the text and performing semantic analysis on the entities. However, this type of method has great limitations. In the process of semantic analysis, the description of semantic information is too mechanized. It is easy to miss, the selection of feature items is one-sided, and the semantic expression is incomplete or even wrong. In short, the accuracy and recall rate of the existing semantic extraction technology cannot meet the requirements, making it difficult to effectively use the text. [0003] Therefore, in the prior art, in the process of tex...

AI Technical Summary

Problems solved by technology

In most information extraction, the semantic information of the text is determined by identifying the entities in the text and performing semantic analysis on the entities, but this type of method has great limitations. In the process of semantic analysis, the description of semantic information is too mechanical. It is easy to miss, the selection of feature items is one-sided, and the semantic expression is incomplete or even wrong. In a word, the accuracy and recall rate of the existing semantic extraction technology cannot meet the requirements, so it is difficult to effectively use the text

Images