[2F6-GS-13] AI application: Emotions and Kansei

[2F6-GS-13-04] A Proposal of an Evaluation Model for Interviewees in Job Interview Scenes Based on Multimodal Information

〇Kento Miyazaki¹, Daisuke Katagami¹ (1. Graduate School of Tokyo Polytechnic University )

Keywords:multimodal information, support vector machine, job interview

In this study, the ability estimation for the job interview scene is carried out. In the field of the human communication, the communication ability estimation using the multimodal information shows the high accuracy, and the similar ability estimation seems to be possible in the job interview which is the interactive field of student and interviewer. In this study, we develop and evaluate a model to estimate the interviewer's evaluation at the job interview using multimodal information such as language features of speech, prosodic features of speech, and motion features such as gesture and posture. For the evaluation of the interviewee, the items of basic skills for working people determined by the Ministry of Economy, Trade and Industry and JAVADA determined by the Ministry of Health, Labour and Welfare are used. Results of evaluation experiments. In the items of "identity" and "first impression", the prosody and action feature sets showed 0.78 estimation accuracies.