A spoken dialogue system that plays the role of an interviewer for job interviews is presented. In this work, our goal is to implement an automated job interview system where candidates can use it as practice before the real inter-view. Conventional job interview systems ask only pre-defined questions, which make the dialogue monotonous and far from human-human interviews. We propose follow-up question generation based on the assessment of candidate responses and keyword extraction. This model was integrated into the dialogue system of the autonomous android ERICA to conduct subject experiments. The proposed job interview system was compared with the baseline system that did not generate any follow-up questions and selected among pre-defined questions. The experimental results show that the proposed system is significantly better in subjective evaluations regarding impressions of job interview practice, the quality of questions, and the presence of the interviewer.