On Modeling of Interviewee Motivation Mental States for an Intelligent Coaching Agent

Abstract

This paper is on agent based model of interview motivation to be integrated in a mental constructs model which serves as a basic mechanics for an intelligent virtual agent coaching for job interview. It has been hypothesized that interview motivation combines with self-efficacy and anxiety to define the mental state of a job interviewee. The concepts were modeled based on psychological theories defining human mental state in a time bounded tasking situation like job interview. The proposed model was formalized and simulated to according to its temporal behaviours. The results of the simulation conform to patterns of a number of relations and casual effects on motivation identified in literature. Additionally, the formal model has been automatically verified using Temporal Trace Language (TTL) to find out which stable situations exist. Consequently, this model can serve as a platform for designing an intelligent agent that can understand the metal state of the user during job interview coaching session.