Research Interests

Advancing Trustworthy Systems for a Smarter World

Research Summary and AI Tools

Mission Statement

The Trustworthy Research in AI and Intelligent Learning Systems Lab (TRAILs) is dedicated to advancing research at the intersection of artificial intelligence, trust, and systems engineering. Our mission is to:

Develop Trustworthy AI Frameworks: Ensure fairness, transparency, and accountability in AI models and applications.
Adapting Ethical AI Practices: Promote responsible AI deployment through global collaboration and interdisciplinary research.
Innovate Intelligent Learning Systems: Build multi-modal AI solutions combining natural language processing, computer vision, and graph analysis for diverse real-world applications.
Address Societal Challenges: Mitigate adversarial threats, detect misinformation, and develop resilient systems for public safety and sustainability.
Empower Communities: Advance AI applications in health, energy, and disaster resilience to positively impact lives.

Research Areas

Social Network Analysis and Trust Modeling

Key Focus Areas
- Modeling trustworthiness and evolution in online social networks.
- Techniques like homophily, preferential attachment, and community detection.
- Offensive content detection in heterogeneous social networks.
Relevant Work
- Trust modeling in online social platforms
- Community-based offensive text detection

Natural Language Processing

Key Focus Areas
- Text normalization for noisy/informal text (e.g., social media).
- Models like BERT, LSTM, and Dual-BERT for sarcasm detection and text analysis.
- Low-resource language processing (e.g., transliteration and summarization for Pashto).
- Defenses against adversarial attacks and explainable AI.
Relevant Work
- Sarcasm detection using Attentive-DualBERT
- Text summarization and dataset development for low-resource languages
- Adversarial attack modeling and defenses on social media platforms

Social/Information Graphs and Behavior Modeling

Key Focus Areas
- Graph neural networks (GNNs) and node embeddings (e.g., Node2Vec, DeepWalk).
- Temporal models (LSTMs, GRUs) and human behavior modeling.
- Link prediction in online social networks.
Relevant Work
- Application of graph-based techniques for trustworthiness analysis
- Activity recognition and modeling during the COVID-19 pandemic
- Friends recommendation

Computer Vision and Multimodal AI

Key Focus Areas
- Vision Transformers (ViTs) and Convolutional Neural Networks (CNNs).
- Multi-task learning for image classification and disease detection.
- Integration of text, image, and graph data for multimodal AI applications.
Relevant Work
- Disease detection using transfer learning
- Face detection in CCTV images for real-time monitoring

Human-Centered and Ethical AI

Key Focus Areas
- Responsible AI and ethical algorithm design.
- AI-driven solutions for societal challenges (e.g., pandemics, misinformation).
- Trustworthiness in AI systems and frameworks.
Relevant Work
- AI-based contact tracing for COVID-19
- Trustworthy AI for social and information ecosystems

Please don't hesitate to contact me if you are interested in collaborating on research in the relevant areas of our joint interest.

Page updated

Google Sites

Report abuse