Automatic matching method, system and storage medium based on subject words and sentence subject matter

Abstract

The present invention provides an automatic matching method, system and storage medium based on subject words and sentence purport, the method includes the following steps: obtaining the first text, extracting the second text from the first text; segmenting the second text, and After segmentation, the text is obtained for encoding, and the encoded sequence information is obtained. According to the encoded sequence information, the first text is bert-encoded, and the information is extracted from the bert-encoded first text to obtain the subject words and gist vectors. According to The subject word and the main idea vector obtain the matching correlation; train the first prediction model according to the subject word, the main idea vector and the matching correlation, and obtain the matching score of the text to be screened and the target text; the method takes up less calculation in the process of execution The resources have stronger content perception capabilities, the amount of data involved in the process is small, and the matching process can be completed more quickly, which can be widely used in the field of natural language processing technology.

Description

technical field [0001] The invention belongs to the technical field of natural language processing, in particular to an automatic matching method, system and storage medium based on subject words and sentence subject matter. Background technique [0002] How to effectively use information search and information matching to enhance social production and communication between people is a problem that many researchers are discussing. Especially in the human resources industry, even though there are many Internet information integration platforms, they are limited by their information matching and technical capabilities, resulting in the emergence of a large number of sunk resumes and recruitment needs. In the recruitment process, enterprises not only have to pay high fees and have low efficiency, but job seekers also present a one-sided job search process. In the recruitment process, the matching of people and positions is the most important application requirement of the huma...

AI Technical Summary

Problems solved by technology

[0005] In addition, using the recommendation system method as the algorithm model of the online recruitment system requires input of a large number of historical recruitment records and resumes, which makes the calculation of the model extremely large, and when the feature extraction ability is insufficient, the matching accuracy will be affected.

Images