Human Centered & Knowledge-infused Learning

In the pursuit of ‘AI for Social Good’, this project builds and apply neuro-symbolic AI approaches (e.g., knowledge-infused learning) across social, clinical, biomedical, and multimodal data. My aim is to design techniques for domain-specific knowledge infusion in ML/DL and NLP at different depths (e.g., shallow, semi-deep, deep infusion). Knowledge infusion mitigates unfairness and socially adverse impacts, reduces false alarms, and covers a broader information space. Currently, I am investigating these mechanisms to learn knowledge-infused representations from social and health data. I apply these techniques in high-impact areas: Cyber Social Threats and Health Informatics.