Seminar: Synthesis of Geometry Proof Problems
Associate Professor, LSU Division of Computer Science and Engineering
Friday September 18, 2020
We present an automated methodology for generating geometric proof problems of the kind found in a high school curriculum. We formalize the notion of a geometry proof problem and describe an algorithm for generating such problems over a user-provided figure. Our experimental results indicate that our problem generation algorithm can effectively generate proof problems in elementary geometry. On a corpus of 110 figures taken from popular geometry textbooks, our system generated an average of about 443 problems per figure in an average time of 4.7 seconds per figure. This is a joint work with C. Alvin (LSU/Furman) S. Gulwani (Microsoft Research, Redmond) and R. Majumdar (Max Planck Institute-SWS).
Dr. Supratik Mukhopadhyay is a faculty member in Computer Science at Louisiana State University. His research interests lie in the areas of Artificial Intelligence/Machine Learning with applications to Education, Automated Drug Discovery, Satellite Imagery Recognition, Transportation Systems, Sustainable Buildings, Cyber-Physical Human Systems, etc. In these areas, Dr. Mukhopadhyay's research has been supported by NSF, NASA, ONR, DARPA, ARO, USDOT, NGA, DOE, NRL, state agencies, and industry. He has more than 110 publications in reputed journals and conferences and has three awarded US Patents. He led the DeepDrug team to the semifinal of the IBM Watson Artificial Intelligence XPRIZE. He is an Associate Editor of IEEE Transactions on Artificial Intelligence and Remote Sensing Letters and a program committee member for AAAI 2021.