AI resume screening method, system, equipment and storage medium

Abstract

The invention discloses an AI resume screening method. The method comprises the steps of inputting plain text contents of all resumes into an improved BERT model, wherein the BERT model adopts a multi-layer Transformer framework; arranging a tutor classifier on the back of the Transformer stack; connecting a student classifier to the back of each layer of Transformer architecture of the BERT model; calculating the classification confidence coefficient of the output result of the student classifier of each layer of Transformer; when the classification confidence coefficient is greater than a preset threshold value, continuing to run the next layer of Transformer, and repeating the previous step; and when the classification confidence is less than the preset threshold value, directly outputting the result. According to the screening method, a recruitment website can simply and quickly perform preliminary screening on massive resume information contents input by job seekers, suspected low-quality resumes are intelligently screened out through a neural network, the calculation speed is high, the model parameter scale is small, and the operability and convenience are provided for next-step accurate screening of the low-quality resumes.

Description

technical field [0001] The invention relates to the field of online recruitment, in particular to a method, system, device and storage medium for intelligently screening resumes by using a neural network model. Background technique [0002] Compared with traditional offline recruitment, online recruitment has great advantages in terms of convenience and information transparency. A general recruitment website or APP client will store a large amount of resume information of job seekers, and with the continuous increase of new job seekers, the resume information of job seekers forms a huge job search data database. Usually, recruitment websites will design different resume templates for job seekers to fill in according to the breadth and level of information required by the enterprise, which largely regulates the resume filling behavior of job seekers. However, in order to highlight the unique experience of job seekers and facilitate the recruiter to fully understand the perso...

AI Technical Summary

Problems solved by technology

This method proposes some solutions for low-quality patents, but it is actually a resume content extraction method, which solves the problems that the resume information extraction process is easily interfered by invalid information, the word segmentation granularity is too small, and the entity reference is unclear. However, There are various types of low-quality patents, and each website has different types of low-quality resumes due to different main directions. This method only focuses on the screening of one type of low-quality resume information, and is not suitable for large online recruitment websites AI resume screening

At the same time, this method uses the neural network model to divide resumes into modules, but it does not optimize the model and training method for low-quality resume screening, and the object of the model output is structured resumes, and there is no special feature for screening low-quality resumes. contribute

[0005] The main purpose of the above-mentioned AI resume screening method is to screen out the main content and core parts of the resume, and does not use technical means to achieve the purpose of filtering low-quality resumes on online recruitment websites

Images