Michael Collins

Michael Collins is the Vikram S. Pandit Professor of Computer Science at Columbia University, where he has been since January 2011. He previously held roles at MIT and AT&T Labs-Research. His research focuses on natural language processing (NLP) and machine learning, emphasizing statistical methods and their applications in parsing, question answering, and cross-lingual systems. Collins teaches courses such as COMS W4705 (Natural Language Processing) and has online teaching materials from a Coursera course in 2013.

Affiliations include Columbia University and Google NYC. He is affiliated with the Department of Computer Science and leads research in NLP. His work spans theoretical foundations like parsing algorithms and practical applications such as machine translation and semantic grounding. Collaborations with students and postdocs have produced influential models and frameworks (e.g., Qed for explainable QA systems). He has also contributed to benchmark datasets like the Natural Questions corpus.

Collins' awards include the NSF Career Award, ACL Fellowship, and multiple best paper awards. His advising includes students like Avner May and Karl Stratos. Research outputs span 20+ years, with recent focus on attribution methods, cross-lingual parsing, and QA systems. His NLP notes and course materials are widely referenced in academic and industrial settings.