Abstract

This research investigates the application of computational linguistic techniques to identify linguistic cues indicative of deception in Arabic political discourse. We analyze a corpus of political speeches and interviews, focusing on features such as sentiment polarity, hedging strategies, lexical diversity, and pragmatic markers. We develop and evaluate a machine learning model trained on these features to detect deceptive statements. The results demonstrate the potential of computational linguistics to uncover subtle linguistic patterns associated with deception in Arabic political communication, offering valuable insights for media analysis, political science, and cross-cultural communication research. The study also addresses the unique challenges of Arabic NLP in the context of deception detection, paving the way for future research in this area.