Prasenjit Mitra
Research Professor, CMU Africa
Affiliated Faculty
Links
Research Areas
Mitra received his Ph.D. in electrical engineering from Stanford University in 2004, his M.S. in computer science from the University of Texas at Austin in 1993, and his B.Tech.(hons.) in computer science and engineering from the Indian Institute of Technology, Kharagpur, India. He worked as a senior member of Technical Staff at Oracle Corporation from 1995-2000. He was one of the co-founders and the inaugural director of the Center for Socially Responsible Artificial Intelligence at the Pennsylvania State University, where he was a tenure-line professor from 2003-2024. From 2018-2021, he served as the Associate Dean for Research for the College of Information Sciences and Technology. He has also supervised or co-supervised over 20 Ph.D. dissertations.
He has co-authored over 200 peer reviewed papers, some of which have been awarded best paper awards, including the Test of Time Award by the IEEE VIS for his paper in the International Conference on Visual Analytics Science and Technology, the premier publication venue in visual analytics in 2021 and a best paper award at ISCRAM, etc. His work has been highly cited (h-index 66, ~16,750 citations). His research has been supported by the NSF CAREER award, the DoE, DoD, NGA, Microsoft Research, Raytheon, Lockheed Martin, Dow Chemicals, McDonnell Foundation, BMBF (Germany), Afretec, etc.
Research Interests
Prasenjit Mitra’s research interests are in the interface of natural language processing and human computer interaction. Specifically, he is interested in information visualization, visual analytics, and their applications usually involving LLMs/foundational models. He is interested in studying the characteristics and behaviour and evolution of LLMs, how human behaviour and digital interactions have changed due to the advent of large foundational models, and how LLMs can be used in applications such as education, health, agriculture, finance, etc. Additionally, the safety and alignment issues of LLMs when it interacts with various types of users and visual analytics using LLMs as an underlying agent are also of interest.