Associate Professor of Information
Associate Professor of Computer Science and Engineering
University of Michigan School of Information and
College of Engineering (affiliate)
Email: kevynct AT umich . edu
School of Information
4341 North Quad
105 S. State Street
Ann Arbor, MI 48109-1285
My lab focuses on developing intelligent information systems that learn when and how to support people's individual information-seeking goals, especially to help people learn and discover. Examples include search engines that can deliver the right kind of personalized information at the right time, and intelligent tutoring systems that learn when and how to be most helpful in teaching a particular student. Building effective, reliable systems like these will require new theoretical, algorithmic, and methodological advances in multiple research areas, including machine learning, optimization, information retrieval, and human-computer interaction. My current application focus is on education, but I'm also very interested in mobile and health-related applications.
One area of special interest is development of robust algorithms for risk-sensitive information systems that can effectively
balance risk and reward, a research direction that I introduced in my PhD thesis. This work connects information retrieval with portfolio theory and other areas of computational finance to arrive at new models,
algorithms, and evaluation methods that account for risk. For example, it shows how the reliability of algorithms for core IR tasks like ranking and query expansion can be greatly improved
by using learning frameworks that jointly optimize for risk and reward objectives.
I'm also interested in large-scale data and text mining, natural language processing, educational applications of IR and machine learning like predicting reading difficulty and computer-assisted language learning, and how the brain acquires language skills.
My Ph.D. is from the School of Computer Science at Carnegie Mellon University, where my advisor was Jamie Callan.
I was a member of the Language Technologies Institute. My undergraduate degree (B.Math.) is from the University of Waterloo. Apparently, I'm not the only one who thinks that CMU and Waterloo are a great combination!
Oct 2014: Happy to be visiting with colleagues this month to give invited talks at CWI in Amsterdam, a workshop on computational learning of language models in Stockholm, and an ETS-sponsored vocabulary assessment workshop in Princeton, NJ.
June 2014: The TREC 2014 Web Track guidelines have been posted here.
Dec 2013: I gave an invited plenary talk on risk-sensitive retrieval in New Delhi at FIRE 2013, the Indian Forum for Information Retrieval Evaluation in New Delhi, Dec 6, 2013.
Dec 2013: I gave an invited talk at the Living Analytics Research Centre Seminar at Singapore Management University on Mon. Dec. 2, 2013.
Nov 2013: The TREC 2013 Web Track has successfully concluded, with some fascinating analysis of the new risk-sensitive retrieval task by several teams. I presented the track overview at TREC 2013: final version of paper to follow shortly.
April 2012: I gave an invited talk for the LTI Colloquium at Carnegie Mellon University on Friday April 27.
April 2012: I gave an invited talk for the CSLP Seminar at Johns Hopkins on Tuesday April 24.
Feb 2012: I enjoyed giving a keynote talk at DDR 2012: Diversity in Document Retrieval, a workshop held on Feb. 12, 2012 in conjunction with WSDM 2012 ACM International Conf. on Web Search and Data Mining. My talk Searching as Investing discussed future research directions that exploit connections with financial optimization to make progress on hard problems in information retrieval.
K. Collins-Thompson. (To appear.) Computational assessment of text readability: a survey of current and future research. ITL International Journal of Applied Linguistics, Sept 2014. (Working paper here)
K. Raman, P.N. Bennett, K. Collins-Thompson. Understanding Intrinsic Diversity in Web Search: Improving Whole-Session Relevance. ACM Transactions on Information Systems (TOIS), Oct. 2014. (pdf)
J. Teevan, K. Collins-Thompson, R. White, S. Dumais. Slow Search. Communications of the ACM 57(8): 36-38 (2014) August 2014. (link)
U.S. Patent 8,700,544. Functionality for Personalizing Search Results. Sontag; David A., Collins-Thompson; Kevyn B., Bennett; Paul N., White; Ryen W., Dumais; Susan T.
K. Collins-Thompson, P. N. Bennett, F. Diaz, C. Clarke, E. Voorhees. TREC 2013 Web Track Overview. NIST Special Publication, Feb 2014. (pdf)
Y. Kim, K. Collins-Thompson, J. Teevan. Crowdsourcing for Robustness in Web Search. NIST Special Publication, Nov. 2013. (pdf)
D. Sontag, K. Collins-Thompson, P. N. Bennett, R. W. White, S. Dumais, B. Billerbeck. Personalization via Probabilistic Adaptation. NIPS 2013 Workshop on Personalization, Dec. 2013. (pdf)
A. Ali, K. Collins-Thompson. Robust Cost-Sensitive Confidence-Weighted Classification. Proceedings of the 8th Workshop on Optimization-Based Techniques for Emerging Data Mining Problems. (OEDM 2013) Dallas, Dec. 2013. (pdf)
J. Teevan, K. Collins-Thompson, R. White, S. Dumais, Y. Kim. Slow Search: Information retrieval without time constraints. Proceedings of HCIR 2013. (pdf)
K. Raman, P.N. Bennett, K. Collins-Thompson. Toward Whole-Session Relevance: Exploring Intrinsic Diversity in Web Search. Proceedings of SIGIR 2013. 463-472. (pdf)[SIGIR Best Student Paper]
F. Raiber, K. Collins-Thompson, O. Kurland. Shame to be Sham: Addressing Content-Based Grey Hat Search Engine Optimization. (Short paper) Proceedings of SIGIR 2013. 1013-1016.
C. Eickhoff, A. de Vries, K. Collins-Thompson. Copulas for Information Retrieval. Proceedings of SIGIR 2013. 663-672. (pdf)
C. Eickhoff, K. Collins-Thompson, P.N. Bennett, S. Dumais. Designing Human-Readable User Profiles for Search Evaluation. Proceedings of ECIR 2013. 701-705. (Short paper) (pdf)
C. Eickhoff, K. Collins-Thompson, P.N. Bennett, S. Dumais. Personalizing Atypical Web Search Sessions. Proceedings of WSDM 2013. 285-294. (Selected for plenary session.)(pdf)
X. Chen, P.N. Bennett, K. Collins-Thompson, E. Horvitz. Pairwise Ranking Aggregation in a Crowdsourced Setting. Proceedings of WSDM 2013. 193-202.(pdf)
K. Collins-Thompson, G. Frishkoff, S. A. Crossley. Definition Response Scoring with Probabilistic Ordinal Regression. Proceedings of ICCE 2012, Singapore, Nov. 2012.(pdf)
L. Wang, P.N. Bennett, K. Collins-Thompson. Robust ranking models via risk-sensitive optimization. Proceedings of SIGIR 2012. (pdf)[SIGIR Best Paper Honorable Mention]
G. Frishkoff, K. Collins-Thompson, C. Perfetti, S. Crossley. Incremental and adaptive word learning from context. SSSR 2012, the Conference of the Society for the Scientific Study of Reading, Montreal, July 2012. Abstract.
J. Kim, K. Collins-Thompson, P. N. Bennett, S. Dumais. Characterizing Web Content, User Interests, and Search Behavior by Reading Level and Topic. Proceedings of WSDM 2012. (pdf)
D. Sontag, K. Collins-Thompson, P. N. Bennett, R. W. White, S. Dumais, B. Billerbeck. Probabilistic Models for Personalizing Web Search. Proceedings of WSDM 2012. (pdf)
K. Collins-Thompson, P. N. Bennett, R. W. White, S. de la Chica, D. Sontag. Personalizing Web Search Results by Reading Level. Proceedings of the Twentieth ACM International Conference on Information and Knowledge Management (CIKM 2011). Glasgow, Scotland. Oct. 2011. (pdf)
P. Kidwell, G. Lebanon, K. Collins-Thompson. Statistical Estimation of Word Acquisition with Application to Readability Prediction. Journal of the American Statistical Association. 106(493):21-30, 2011. (pdf)
K. Collins-Thompson. "Improving information retrieval with reading level prediction." SIGIR 2011 Workshop on Enriching Information Retrieval. Beijing, July 2011. (pdf)
G. Frishkoff, C. Perfetti, and K. Collins-Thompson. "Predicting robust vocabulary growth from measures of incremental learning". Scientific Studies of Reading, 15(1), 71-91. January 2011.
J. Dillon and K. Collins-Thompson. A unified optimization framework for robust pseudo-relevance feedback algorithms. Proceedings of the Nineteenth ACM International Conference on Information and Knowledge Management (CIKM 2010), Toronto, Canada. (CIKM Student Travel Award Paper) (pdf)
M. Heilman, K. Collins-Thompson, M. Eskenazi, A. Juffs, L. Wilson. "Personalization of reading passages improves vocabulary acquisition." International Journal of Artificial Intelligence in Education, 20(1), 2010. (pdf)
J. Huang, N. Koudas, G. Jones, X. Wu, K. Collins-Thompson, and A. An. (eds.) Proceedings of the Nineteenth ACM International Conference on Information and Knowledge Management (CIKM 2010), ACM Press, New York.
K. Collins-Thompson and J. Dillon. Controlling the search for expanded query representations by constrained optimization in latent variable space. SIGIR 2010 Workshop on Query Representation and Understanding.(pdf)
Frishkoff, G. A., Perfetti, C. A., & Collins-Thompson, K. (2010). Lexical quality in the brain: ERP evidence for robust word learning from
context. Developmental Neuropsychology, 35(4), 1-28. [details]
M. Sun, G. Lebanon, and K. Collins-Thompson. "Visualizing Differences in Web Search Algorithms using the Expected Weighted Hoeffding Distance". Proceedings of WWW 2010, Raleigh, NC, U.S.A. pg 931-940. (pdf)  [bibtex]
K. Collins-Thompson, P.N. Bennett. "Predicting query performance via classification", Proceedings of ECIR 2010, Milton Keynes, UK. pg 140-152. (pdf)  [bibtex]
K. Collins-Thompson. "Reducing the risk of query expansion via robust constrained optimization". Proceedings of the Eighteenth International Conference on Information and Knowledge Management (CIKM 2009). ACM. Hong Kong. pg. 837-846.(pdf)  [bibtex]
K. Collins-Thompson. "Accounting for stability of retrieval algorithms using risk-reward curves". Proceedings of SIGIR 2009 Workshop on the Future of Evaluation in Information Retrieval, Boston. pg. 27-28.(pdf)
M. Sun, G. Lebanon, and K. Collins-Thompson. Visualizing Spatial Proximity of Search Algorithms, NIPS Workshop on Learning with Ordering. (Poster abstract), 2009. (pdf)
K. Collins-Thompson. "Robust word similarity estimation using perturbation kernels". Proceedings of the International Conference on Theoretical Information Retrieval (ICTIR) 2009, Cambridge, U.K. pg. 265-272.(pdf)  [bibtex]
P. Kidwell, G. Lebanon, K. Collins-Thompson. "Statistical estimation of word acquisition with application to readability prediction". Proceedings of Empirical Methods in Natural Language Processing (EMNLP) 2009, Singapore. (pdf)
K. Collins-Thompson, P. N. Bennett. "Estimating query performance using class predictions". Proceedings of the Thirty-second Annual International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR 2009), Boston. pg. 672-673. (Poster description) (pdf)  [bibtex]
K. Collins-Thompson. "Estimating robust query models with convex optimization". Advances in Neural Information Processing Systems 21 (NIPS), 2008. pg. 329-336.(pdf)   [bibtex]
K. Collins-Thompson. "Robust model estimation methods for information retrieval". Ph.D. thesis (LTI Technical Report CMU-LTI-08-010) Carnegie Mellon University, 2008.
G. Frishkoff, K. Collins-Thompson, C. Perfetti, J. Callan. Measuring incremental changes in word knowledge: Experimental validation and implications for learning and assessment. Behavior Research Methods, Vol. 40, No. 4. pp. 907-925. (pdf)  [pubmed]
M. Heilman, K. Collins-Thompson and M. Eskenazi. "An analysis of statistical models and features for reading difficulty prediction." ACL 2008 BEA Workshop on Innovative Use of NLP for Building Educational Applications. Columbus, Ohio. (pdf)
K. Collins-Thompson and J. Callan. "Estimation and use of uncertainty in pseudo-relevance feedback." Proceedings of the Thirtieth Annual International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR 2007), Amsterdam. (pdf)  [bibtex]
K. Collins-Thompson and J. Callan. "Automatic and human scoring of word definition responses." Proceedings of the NAACL-HLT 2007 Conference. Rochester, U.S.A. pp. 476-483. (pdf)  [bibtex]
K. Collins-Thompson. Optimization methods for query model estimation: applying portfolio theory to mitigate risk in information retrieval. CMU DIR Group Technical Report 2007-09-03. Abstract
M. Heilman, K. Collins-Thompson, J. Callan and M. Eskenazi. "Combining lexical and grammatical features to improve readability measures for first and second language texts." Proceedings of the NAACL-HLT 2007 Conference. Rochester, U.S.A. pp. 460-467. (pdf)  [bibtex]
M. Heilman, K. Collins-Thompson, J. Callan, and M. Eskenazi. Classroom success of an Intelligent Tutoring System for lexical practice and reading comprehension. Proceedings of Interspeech 2006. Pittsburgh, U.S.A. abstract
A. Juffs, L. Wilson, M. Eskenazi, J. Callan, J. Brown, K. Collins-Thompson, M. Heilman, T. Pelletreau, and J. Sanders. (2006) "Robust learning of vocabulary: investigating the relationship between learner behaviour and the acquisition of vocabulary" (poster). The 40th Annual TESOL Convention and Exhibit (TESOL 2006).
K. Collins-Thompson and J. Callan. Query expansion using random walk models. Proceedings of the Fourteenth International Conference on Information and Knowledge Management (CIKM'05). ACM. Bremen, Germany. (CIKM Student Travel Award Paper) (pdf)  [bibtex]
K. Collins-Thompson, J. Callan. Predicting reading difficulty with statistical language models. Journal of the American Society for Information Science and Technology. Vol. 56, No. 13, 1448-1462.  [bibtex]
K. Collins-Thompson, P. Ogilvie and J. Callan. Initial results with structured queries and language models on half a terabyte of text. Proceedings of TREC 2004, National Institute of Standards and Technology, special publication. (pdf)
K. Collins-Thompson and J. Callan. A language modeling approach to predicting reading difficulty. Proceedings of HLT / NAACL 2004, Boston, USA, May 2004. (pdf)  [bibtex]
K. Collins-Thompson and J. Callan. Information retrieval for language tutoring: an overview of the REAP project (poster description), Proceedings of SIGIR 2004, Sheffield, UK. July 2004. (pdf)  [bibtex]
K. Collins-Thompson, E. Terra, J. Callan, and C. Clarke. The effect of document retrieval quality on factoid question-answering performance (poster description), Proceedings of SIGIR 2004, Sheffield, UK. July 2004. (pdf)  [bibtex]
J. Zhang, A. Toth, K. Collins-Thompson, and A. Black. Prominence prediction for super-sentential prosodic modeling based on a new database, ISCA Synthesis Workshop, Pittsburgh, USA, June 2004.
E. Nyberg, T. Mitamura, J. Callan, J. Carbonell, R. Frederking, K. Collins-Thompson, L. Hiyakumoto, Y. Huang, C. Huttenhower, S. Judy, J. Ko, A. Kupsc, L. V. Lita, V. Pedro, D. Svoboda, and B. Van Durme. (2004.) "The JAVELIN question-answering system at TREC 2003: A multi-strategy approach with dynamic planning." Proceedings of the 2003 Text REtrieval Conference (TREC 2003). National Institute of Standards and Technology, special publication. (pdf)
U.S. Patent 6,735,335. M. Liu, K. Collins-Thompson, D. Lawton. Method and apparatus for discriminating between documents in batch scanned document files. May 2004.
U.S. Patent 6,687,697. K. Collins-Thompson, C. Schweizer. System and method for improved string matching under noisy channel conditions. Feb. 2004.
K. Collins-Thompson, P. Ogilvie, Y. Zhang, and J. Callan. Information filtering, novelty detection, and named-page finding. In Proceedings of the 2002 Text REtrieval Conference (TREC 2002). National Institute of Standards and Technology, special publication. 107 - 118.(pdf)
E. Nyberg, T. Mitamura, J. Carbonell, J. Callan, K. Collins-Thompson, K. Czuba, M. Duggan, L. Hiyakumoto, N. Hu, Y. Huang, J. Ko, L. Lita, S. Murtagh, V. Pedro, D. Svoboda. The JAVELIN Question-Answering System. In Proceedings of TREC 2002. NIST, special publication. 128 - 137.
Reviewer, ACM Transactions on Information Systems; ACM Transactions on the Web; IEEE Transactions on Knowledge and Data Engineering; Information Processing and Management; Foundations and Trends in Information Retrieval; Transactions on Audio, Speech, and Language Processing; Journal of the American Society for Information Science and Technology.
WSDM 2013 Crowdsourced Pairwise Preferences for Readability (.csv file, 9.1Mb): 13857 judged pairs (trusted and untrusted), ~50-word text passages, grades 1-12. Column descriptions are here.
If you use this dataset, please cite: X. Chen, P.N. Bennett, K. Collins-Thompson, E. Horvitz. Pairwise Ranking Aggregation in a Crowdsourced Setting. Proceedings of WSDM 2013. 193-202.