Development of a Web-Based Interview Support System Using Characters Nodding with Various Movements

Abstract

In contrast to conventional face-to-face interviews, web-based interviews may cause additional tension due to the lack of a sense of unity and difficulty in sharing the rhythm of dialogue, in addition to the tension of job interviews. First, we conducted a questionnaire survey on job interviews to find out what kind of interview format and environment actually reduces nervousness. The questionnaire included questions about interview methods and emotions during interviews for fourth-year university students who had experienced job interviews. The results showed that many of the respondents felt that the explanations and interviews before the interviews helped to alleviate their anxiety and nervousness. In addition, many respondents said that they had casual conversations with their friends and interviewers before the interview to ease their tension. Furthermore, the presence of a familiar person at the interview or not may have helped to alleviate the nervousness of the interviewees. Based on these results, we propose a web interview support system using the communication support character called InterActor which is a speech-driven embodied entrainment character that automatically generates body movements based on speech. We developed a web interview support system using the nodding patterns grouped by cluster analysis based on the impressions. The effectiveness of the developed system was demonstrated in an evaluation experiment using a role-play interview task.