DAC Student Spotlight: Debanjan Datta
Debanjan Datta’s interest in data mining focuses on systems that perform anomaly detection with both interpretability and the ability to incorporate domain knowledge and human input.
In a recent Discovery Analytics Center study with the World Wildlife Fund, Datta developed a framework that can apply machine learning on massive trade datasets to detect patterns of suspicious timber records that relate to possible illegal trade. He shared results of the study, “Detecting Suspicious Timber Trades,” at the Conference on Innovative Applications of Artificial Intelligence (IAAI) earlier this month.
The research involves analyzing, record-by-record, thousands of lines of export and import data.
“By analyzing available timber data, along with open source domain knowledge, we are trying to develop software and algorithms that will help flag suspicious timber at the border in real time. Such a human-machine approach can improve both efficiency and effectiveness,” Datta said.
Datta, a Ph.D. student majoring in computer science, is advised by Naren Ramakrishnan.
“Projects aimed at solving real world challenges with state of the art approaches that are not restricted to one area really piqued my interest in joining DAC,” said Datta. “DAC’s approach to research offers a breadth that is difficult to find.”
Flexibility to explore research areas and opportunities to collaborate on interesting projects and learn from people having a myriad of interests are major positive aspects of being a DAC student, he said.
Datta earned a bachelor of engineering in computer science and engineering from Jadavpur University, India, and was a software development engineer at Yahoo, Inc. both in Bangalore, India, and in Sunnyvale, California, prior to pursuing his Ph.D.
He is projected to graduate in Fall 2021. After graduation, Datta said he “would like to continue in research, probably in a lab that carries on work with practical applications.”