RESEARCH THEMES

 Methodological advances in sentiment analysis, sarcasm detection, stance detection, deception and misinformation detection, low-resource language processing, and LLM evaluation in high-impact domains.   

Theoretical and applied research on model interpretability, algorithmic accountability, risk mitigation, and governance frameworks for high-stakes AI systems.  

 Modelling online influence, extremism, hate-speech, racism, misogyny, cyber resilience, and algorithmic amplification using network science and socio-technical frameworks.  

Harnessing multimodal data, including IoMT bio-signals, clinical text, and social media cues for early disease prediction, emotion recognition and detection, stress and pain detection, mental health classification and privacy-preserving health AI systems.