Interview Bot Development with Natural Language Processing and Machine Learning

Interview for competency assessment takes essential roles in Human Resource Management practices. However, the traditional competency interview process needs considerable time and costs and often requires face-to-face meetings that may endanger both interviewers and interviewees during a pandemic. This study aims to present the development of an interview bot for identifying competency based on the Behavioural Event Interview method by using artificial intelligence technology. It is an automation of the interview process to explore a person's competencies levels based on past behavioural experiences. The development of the interview bot involved two main activities. The first is the data training process to develop learning models to determine competency levels based on provided valid participant's responses. The second is the testing and evaluation model for assessment to determine competency levels. We found that our method can predict a person's competence levels based on their responses. Our approach can make predictions with acceptable accuracy. The interview bot is a valuable and reliable tool to conduct online interviews and support the assessment centre process, especially with conditions of physical and social distancing constraints. It provides flexibility in terms of time and place for participants, and its process is delivered in Indonesia's Language. The interview bot is more cost efficient than traditional interviews with the same behavioural event interview methods, and it would also be preferable for millennials.