David Jurgens

David Jurgens is an associate professor jointly in the School of Information and the Department of Electrical Engineering and Computer Science at the University of Michigan. He holds a PhD from the University of California Los Angeles and was a postdoctoral scholar in the Department of Computer Science at Stanford University and prior at McGill University. His research combines natural language processing, social psychology, and data science to discover, explain and predict human behavior in large social systems. His research has been published in top computational social science and natural language processing venues including PNAS, WWW, ACL, ICWSM, EMNLP, and others. His work has won the Cozzarelli Prize from the National Academy of Science, Cialdini Prize from the Society for Personality and Social Psychology, best paper at ICWSM and W-NUT, best paper nomination at ACL and Web Science, and has been featured in news outlets such as the BBC, Time, MIT Technology Review, New Scientist, and Forbes.

How he got there: Before joining UMSI, David was a postdoctoral scholar, jointly in the Stanford NLP and SNAP Groups under Dan Jurafsky, Jure Leskovec and Dan McFarland. Prior, he ventured beyond the wall to the cold regions of Montreal (don't let the idyllic summers fool you) and was a postdoctoral scholar at McGill University in the Network Dynamics group with Derek Ruths. Before finishing his PhD, he was a research scientist at the Linguistics Computing Laboratory at Sapienza University of Rome under Roberto Navigli. During his PhD, he was concurrently a visiting researcher at the Information and Systems Science Lab at HRL Laboratories. After trips abroad and to Malibu, he received his PhD in Computer Science from the University of California, Los Angeles under Michael Dyer. Early in his career before he discovered you could study language and people, he received his BA in Philosophy and Political Science and an MS in Computer Science on Computer Vision under Robert Pless from Washington University in St. Louis.

Current

Winter 2024, Fall 2025 — EECS 595, Natural Language Processing

Winter 2026 — SI 405, Applied Generative AI

Winter 2026 — CSE 598-004, Building Small Language Models

Fall 2025 — SI 670, Applied Machine Learning

Past

Fall 2017, 2018 — SI 671, Data Mining: Methods and Applications

Winter 2017&ndash 2024; SI 630, Natural Language Processing: Algorithms and People

Fall 2020–2023 — SI 650 / EECS 549, Information Retrieval

Winter 2023 — SI 330, Data Manipulation

Fall 2019 — SI 710, PhD Seminar: Computational Sociolinguistics
A new course, on computational sociolinguistics — with actual computation and actual sociolinguistics. If any of that excites you, drop me an email and I can share more details.

Being a faculty involves too many things to reasonably keep track of, so this page is just left out of date in favor of staying sane. For folks who are very interested still, please see my CV for a more up-to-date list.

Current

• Co-editor of a Frontiers special issue on Computational Sociolinguistics. The journal has a rolling deadline so feel free to submit here, or just read the papers as they appear.
• Area chair for Social Media for ACL and EMNLP.
• Sponsorship chair for ICWSM 2019.
• Senior PC for WWW 2020 (Web & Society).
• Co-chair of the International Workshop on NLP and Computational Social Science at ACL-2017 with Dirk Hovy, David Bamman, Oren Tsur, and Svitlana Volkova.

Past

• Data Chair for ICWSM 2017.
• Co-chair of the International Workshop on NLP and Computational Social Science at ACL-2017 with Dirk Hovy, David Bamman, A. Seza Dogruoz, Brendan O'Connor, Oren Tsur, and Svitlana Volkova.
• Co-chair of the International Workshops on NLP and Computational Social Science at EMNLP-2016 and at WebSci-2016 with Dirk Hovy, David Bamman, A. Seza Dogruoz, Jacob Eisenstein, Brendan O'Connor, Alice Oh, Oren Tsur, and Svitlana Volkova.
• General reviewing: Program Committee at various (most) times for WWW, ICWSM, NAACL, ACL, EMNLP, CSCW, LREC, EACL; External Reviewer for CHI (rarely).
• Co-chair of the International Workshop of Semantic Evaluation (SemEval) in 2015 with Daniel Cer, Preslav Nakov, and Torsten Zesch and in 2016 with Daniel Cer, Marine Carpuat, and Steven Bethard.
• Co-organizing SemEval-2016 Task 14: Semantic Taxonomy Enrichment with Mohammad Taher Pilehvar.
• Co-presented the tutorial “Semantic Similarity Frontiers: From Concepts to Documents” with Mohammad Taher Pilehvar at EMNLP-2015 — over 150 registered attendees. See our annotated bibliography.
• Co-presented the tutorial “Multilingual Semantic Processing” with Roberto Navigli at LREC-2014.
• Co-organizer of SemEval-2014 Task 3: Cross-Level Semantic Similarity with Mohammad Taher Pilehvar and Roberto Navigli.
• Co-organizer of SemEval-2013 Task 12: Multilingual Word Sense Disambiguation with Roberto Navigli and Daniele Vannella.
• Co-organizer of SemEval-2013 Task 13: Word Sense Induction for Graded and Non-Graded Senses with Ioannis Klapaftis.
• Co-organizer of SemEval-2012 Task 2: Measuring Degrees of Relational Similarity with Saif M. Mohammad, Peter D. Turney, and Keith J. Holyoak.
• Journal Reviewer: Artificial Intelligence, Behavior Research Methods, International Journal of Corpus Linguistics, Language Resources and Evaluation, Natural Language Engineering, PLOS ONE, Transactions on Knowledge Discovery from Data (TKDD).

Broadly, I conduct research in natural language processing. My current work focuses on three central themes.

• Social Reasoning. Social settings are complex. We study how people reason about social situations, and how language and behavior change in social contexts. Our work is grounded in social and cognitive psychology, and develops new computational methods to study language and mental models.
• Human–AI Collaboration in Evaluation. Most NLP models are designed to do one or more tasks. To train or assess how good those models are, we need ground truth — which can be very challenging to create. We examine how and when we can use humans and AI systems together, or individually, to better evaluate NLP models even for the most complex tasks.
• Information Ecosystems. The interconnected and rapid nature of news and social media means people can get new information almost anywhere, anytime. How does news spread, and who does it reach, especially as it crosses social, linguistic, or medium boundaries? We study whole ecosystems of how the language of information changes and the social processes by which it emerges and evolves.

My long-term research goal combines human and language technologies to create social understanding that reflects both the content and the people involved in communication. In all my research, I strive to improve social equality by representing all the people participating in these social systems.

Masters and Undergraduate Students

For current students, due to the technical work we do in the lab, I typically require students to have taken some class on NLP or advanced machine learning to give them the requisite skills. Without those classes we end up teaching the same techniques in a less principled way, which takes more time. If you feel like you already have significant research experience (just not in NLP or ML), please explain what you've done. Students are typically expected to join our group's research meetings. During the school year we'll typically have one meeting a week that is also with the project's co-supervisor (a PhD student or postdoc). Interested students should apply through this form.

PhD Students

I admit roughly one PhD student per year. Sometimes students are co-admitted or co-advised, so the number of admits can vary.

For prospective PhDs, I especially like students who come with a strong computational background and some experience in social science. There are no set criteria, but you're much better off in admissions if you've contacted me (or your advisor has) and let me know of your interests and goals. Make sure to look over these pages carefully; the match should be pretty strong. A PhD student is very costly — in time and money — and I select students for my research group carefully.

If you're a current PhD student outside of CSE or SI, I'm open to collaborations. One of the best things about SI and CSE are the interdisciplinary environments, and I'm potentially open to hosting students outside my home departments (but inside UM) in lab or co-advising on projects where it makes sense. Regardless, I'd love to hear from you, and you're always welcome to come take my classes.

PhD Students not at the University of Michigan

If you're a PhD student somewhere else and want to work with me (while being external), this could happen under the right circumstances. Typically, your advisor at your primary institution and I would co-advise you on a specific project. I usually only do these kinds of arrangements when I know your advisor (more common) or when the collaborative project makes sense (rare). To get this started, have your advisor email me (not you directly) about what the project is.

Non-PhD External Students

I unfortunately rarely work with high-schoolers, undergraduates, or master's students who are not physically at the University of Michigan. I still get emails from external students asking if we could collaborate remotely, and I really would love to, but my priority is to advise the current students at UM given the limited bandwidth I have for advising. Your best bet to work with me is to get admitted to one of our programs and then drop me an email.

Postdocs

I would love to have you all in my lab, but this is generally dependent on funding (but seriously, I would take you all if I could). Email me if you think you're a good match and tell me why, and we might be able to figure something out. At the moment I'm not actively seeking postdocs (due to funding, of course). If you're coming with your own funding, that changes everything — drop me a line.

How do I pronounce your last name?

Like you would in the old country.

Which old country is that?

🤷

Can I research with you or be a member of the Blablablab?

For an overly-detailed answer, click the “Students” tab above. That should cover everything.