An Innovative Approach of Personality Recognition for E-Recruitment

Abstract

The field of online recruitment systems is becoming more popular in Artificial Intelligence because it is beneficial for both candidates and interviewers as it saves time and energy. In the manual process of recruitment, fitting the job specifications according to the resume and selecting the perfect candidate as per their behavior is a difficult task. With uses in psychiatric evaluations, human operator, and personality computing, automatic analysis of video interviews and automatic extraction of resumes for recognizing personality traits has consequently emerged as an important research subject. Convolutional neural network (CNN) models were introduced in some earlier studies as a result of developments in Deep Learning (DL)-based computer vision and pattern recognition. These models are capable of accurately predicting human non-verbal cues when used in conjunction with a web camera. In this paper, the candidate and interviewer both can achieve their goals by the one system. As per job specification included in the resume, candidates can get clarification of the job title and test their own personality by giving a psychometric assessment included in the system. The end-to-end AI interviewing system is developed with the aid of asynchronous video interview (AVI) processing, and automatic personality identification (APR) is carried out using features gleaned from the AVIs by the Tensorflow AI engine. The result shows that the interviewer can successfully recognize the Big five personality traits of a candidate at an accuracy above 95%. In the automatic personality recognition the semi supervised DL approach gives better performance.