Unveiling Candidates’ Traits and KSA Mapping Through Video Analysis in Interviews

Recruitment today in its dynamic scenario gets more involved with non-technical indicative features like personality attributes or interpersonal skills since it is all important in the present. This particular research focuses on an AI-based framework that is focused on the use of asynchronous video interviews by means of predicting the Big Five personality traits of candidates with respect to Knowledge, Skills, and Abilities (KSAs), which are otherwise highly relevant to its roles. Convolutional Neural Networks (CNNs) were used for feature extraction followed by rule-based mapping, which has resulted in highly accurate predictions, for example, MSE 0.15 levels of prediction for conscientiousness and MSE 0.23 for openness. The mappings of the predicted traits with KSAs, such as a conscientiousness-responsibility link, were validated by industry experts. This study addresses some limitations, one being the restriction of features extracted from the dataset. It also emphasizes ethical recruitment based on AI systems. The framework has the potential to make hiring more objective and fair while providing a scalable solution in this direction within recruitment.