“Big data” will not change the world unless it is collected and synthesized into tools that have a public benefit.
Our research is organised around three interconnected pillars:
We develop novel methods in machine learning, artificial intelligence, statistical inference, and computational social science to understand complex systems and support better decision-making.
Technology should be designed with people, not simply for people. We integrate human-computer interaction, participatory design, and behavioral science to ensure systems are usable, trustworthy, inclusive, and aligned with human values.
We harness AI and data science to support evidence-informed decision-making in governments, international organizations, and civil society. Our work spans public policy, digital governance, health, education, energy, and sustainability.
100+ substation nodes monitored, advising power distribution policy.
Real-time GDP mapping methodology via grid load indicators.
National canal/river freight path navigation planning engine.
Smart infrastructure & climate resilience.
Dr. Moinul Islam Zaber presented on bridging technical AI ethics with regulatory enforcement to balance societal innovation with legal boundaries.
Collaborative project integrating immersive interactive AI exhibits and VR headsets to bring history education alive for high school students.
Panel discussing methodologies to make black-box deep learning models transparent for medical, regulatory, and public welfare sectors.
Dr. Zaber spoke on 'Data Materialities, Infrastructures, Critiques' discussing the physical constraints and ecological footprint of computational pipelines.
Authors: Murshed Al Amin, MD. Jawad-Al-Mursalin Hoque, Zakia Nazzum, Md Abu Sayed, Sarker Tanveer Ahmed Rumee, Moinul Islam Zaber
Proposes a robust handwriting recognition pipeline to digitize raw logs from remote electricity substations, assessing sensor quality errors.
Authors: Shamim Bin Zahid, Aishwarya Ghosh Bristy, Sarker Tanveer Ahmed Rumee, Moinul Islam Zaber
Applies NLP techniques to summarize and flag predatory privacy clauses hidden inside lengthy Software End-User License Agreements.
Authors: Firoza Tabassum, Humaira Islam, Matin Saad, Moinul Islam Zaber
Applies graph-theoretical models to AIS transponder coordinates to discover connectivity bottlenecks across delta shipping lines.
Authors: Md. Mehedi Hasan, Sarker Tanveer Ahmed Rumee, Moinul Islam Zaber
Establishes a localized dataset and deep convolutional-recurrent model to class acoustic emotional features in native Bangla dialect.
Something else entirely? Write to us anyway.
We look for research collaborations, corporate partnerships, and brilliant prospective graduate students to push the boundaries of data science and HCI.
Email: zaber@du.ac.bd
Office: Room 402, Dept of CSE, University of Dhaka
