SYSTEM FOR EMOTION CLASSIFICATION IN INTERVIEW SETTINGS

Abstract

The traditional hiring process can be time-consuming and expensive for companies, often requiring multiple interviews and lacking objectivity. This project introduces Emotico, a web application that streamlines and enhances the hiring process. Emotico allows recruiters to post job openings and associated interview questions. Candidates then take these interviews online, with their responses evaluated by a combination of advanced technologies. Emotico leverages OpenAI's ChatGPT-4 model to analyze the content of a candidate's answers, categorizing them and providing a score with an explanation. Additionally, Emotico incorporates emotion detection through a multimodal CNN architecture, developed by Chumachenko et al. (2022), to analyze emotions both from video and audio during video interviews. This approach provides a more well-rounded assessment of a candidate's suitability for the role. Emotico offers significant advantages over traditional hiring methods. It streamlines the process by allowing companies to conduct interviews online and receive automated evaluations. The combination of text analysis and emotion detection offers a more comprehensive understanding of candidates, potentially reducing bias and leading to better hiring decisions. Emotico's development can be further enhanced by incorporating additional features and refining the Machine Learning models. Exploring new avenues for emotion detection and integrating with applicant tracking systems are promising areas for future exploration.