DEVELOPMENT OF A NOVEL METHOD FOR DETECTION OF FAKE VS HONEST HUMAN BEHAVIOUR USING MACHINE LEARNING TECHNIQUES

Abstract

Human behaviour analysis is increasingly vital across security, recruitment, mental health, and cybersecurity domains. This study proposes a novel machine learning (ML) framework to classify human behaviour as honest or deceptive by leveraging multimodal data, including facial expressions, audio cues, and textual content. The system extracts salient behavioural features and employs both classical and deep learning models for detection. Experimental results demonstrate that a multimodal fusion approach significantly outperforms single-modality models, offering a robust, scalable, and non-invasive solution for real-time authenticity assessment.