Chat Lab News

Abm Adnan Azmee successfully defended his doctoral dissertation, titled “Advancing Human-AI Collaboration for Behavioral and Mental Health Identification.” His research focuses on developing scalable, interpretable, and adaptive human–AI teaming frameworks to analyze unstructured data. The work integrates domain knowledge, multi-level attention mechanisms, and human-in-the-loop feedback to improve early detection of behavioral and mental health conditions. This dissertation contributes to advancing intelligent systems that support real-world decision-making in safety-critical and healthcare settings.

Abm Adnan Azmee was honored with the Doctoral Dissertation Award from the KSU Graduate College in recognition of the originality, technical depth, and societal impact of his research. The award highlights dissertations that demonstrate exceptional scholarly contribution and innovation. His work stood out for developing novel Human-AI frameworks to address pressing challenges in behavioral and mental health identification, particularly in real-world, high-impact environments.

Abm Adnan Azmee received the Outstanding PhD Student Award 2026 from the College of Computing and Software Engineering at Kennesaw State University. This recognition highlights his strong record of doctoral research, scholarly publication, and interdisciplinary collaboration in Human-AI collaboration, natural language processing, and behavioral health computing. His work developed novel AI-driven methods that integrate expert knowledge, explainability, and human feedback to support behavioral and mental health identification from real-world data, demonstrating both technical innovation and meaningful societal impact.

"I successfully presented my PhD dissertation proposal, "Developing Scalable and Collaborative Explainable Human-AI Systems for Behavioral Health Analysis". This research project tackles a major challenge at the intersection of explainable AI (XAI), human-AI collaboration, and behavioral health: how to use AI to detect behavioral health signals in sensitive data while maintaining transparency and trust among practitioners making high-stakes intervention decisions.

My proposed approach combines human-in-the-loop active learning and explainable AI techniques, enabling domain experts to contribute on refining domain-grounded explanations. This work integrates machine learning, natural language processing, and human-computer interaction to create systems that improve, not replace, human judgment.

I am grateful to my advisor, Dr. Md Abdullah Al Hafiz Khan, and my committee, Drs. Yong Pei, Monica Nandan, and Kazi Islam, for their insightful feedback and support. Cheers!"

Abm Adnan Azmee presented his work on Human-AI collaboration at IEEE/ACM CHASE 2025, a leading venue in connected health research. CHASE brings together interdisciplinary work spanning sensing, communications, and intelligent analytics for health applications and systems. The 2025 conference was hosted in Manhattan, New York City, USA (June 24–26, 2025), featuring a multi-day program of keynotes and technical sessions.

Kennesaw State University’s College of Computing and Software Engineering honored Francis Nweke with the Outstanding Ph.D. Student Teaching Award (2025) for excellence in teaching. This award acknowledges exceptional instructional contributions and dedication to student success within the PhD in Computer Science program.

Abdul Muntakim presented at IEEE ICMLA 2025. This work introduces a novel Trust-Aware Human-AI Teaming (HAT) framework for reliable text classification in high-risk and ambiguous scenarios. The approach integrates synthetic cognitive factors—including trust in model advice, cognitive load, perceived difficulty, probabilistic correctness, and historical accuracy—into decision-making using outputs from multiple large language models.

A gated human–AI mechanism dynamically balances automated predictions and human review, activating human oversight only when confidence or trust is low. Evaluated on over 15,000 news articles annotated by four state-of-the-art LLMs, the proposed model achieves 82.8% accuracy and 82.1% F1 score, outperforming standalone and static ensemble baselines while requiring human review for only 39.47% of cases.
Venue: Published in IEEE International Conference on Machine Learning and Applications (ICMLA), 2025.

Abm Adnan Azmee presented his work on multi-label behavioral health identification from police narrative reports at IEEE ICMLA 2024. He also served as a Session Chair at the event. ICMLA serves as a leading international forum for disseminating original machine-learning research, with emphasis on real-world applications and novel systems. The 2024 event took place at Coral Gables, Florida, USA (Dec 18–20, 2024), supporting both research exchange and practitioner engagement.

Francis Nweke presented his research titled “Explainable Multilabel Classification Framework for Behavioral Health Based on Domain Concepts” at the IEEE International Conference on Big Data (BigData 2024), held in Washington, D.C., USA (December 15–18, 2024). The work highlights how domain-driven behavioral health concepts improve interpretability in multilabel classification and how annotated 911 narratives can advance explainable AI for public health.

Abm Adnan Azmee presented research on his developed domain-enhanced attention network for behavioral health identification from 911 narratives at IEEE BigData 2023. IEEE BigData is an international conference that convenes researchers and practitioners across big-data methods, systems, and applications. The 2023 conference was held in Sorrento, Italy (Dec 15–18, 2023), including special sessions focused on data mining and machine learning on big data.

Kennesaw State University’s College of Computing and Software Engineering recognized Abm Adnan Azmee with the Outstanding Ph.D Student Research Award (2024) for research excellence. The honor reflects sustained scholarly contributions and research impact within the PhD in Computer Science program.