Case extraction method based on natural semantic analysis

Abstract

The invention belongs to the field of data processing technology and discloses a case extraction method based on natural semantic analysis. The case extraction method comprises the steps that a word bank belonging to a self-domain is trained through a large number of relevant professional demand documents; the precision of the word bank is optimized, and mistaken vocabularies disturbing a word segmentation result are eliminated; a corpus of a relevant domain is adopted to perform word segmentation, and cases and participants are identified; a sentence trunk is extracted on the basis of word segmentation, and a semantic tree based on semantic analysis is established; and the cases are recognized and extracted through model matching according to the extracted sentence trunk. According to a new word discovery algorithm, three characteristics are utilized to well extract needed new words; in the machine learning process, a lot of parameter control and the like are performed, meanwhile, a lot of sample learning is performed, five types of Chinese grammar models are generalized, therefore, case extraction and participant matching are realized, and the final accuracy is determined at 95% or above.

Description

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine

Login to view more

AI Technical Summary

Benefits of technology

This patented method improves how data from different sources match with each other better when used correctly across various applications such as communication or computer science systems. It achieves these technical benefits by utilizing techniques like fuzzy logic and neural networks to learn complex relationships based on user-generated content (UGC) inputted into an artificial intelligence engine called Sage). These methods help us create customized documents quickly without losing important parts during processing time due to lack of precision matches.

Problems solved by technology

Technologies aim towards improving our understanding about complex systems involving multiple aspects - including domain boundaries, linguisms, symbolism, and sensory relationships among them. These advancestances help us better interpret speeches through mathematical principles rather than just categorizing things differently.

Images