Jason D. M. Rennie
44 Grove St.
Belmont, MA 02478
jrennie@gmail.com
Work/Research Interests
- Attacking real-world problems which apply my knowledge of
preference learning and collaborative filtering, especially where a
multiplicity of factors combine to produce a discrete-valued output
- Automated data analysis for classification, estimation and
acquisition of new knowledge
Education
Massachusetts Institute of Technology, Cambridge, MA
Ph.D. in Electrical Engineering and Computer Science, January 2007
Thesis Title: Extracting Information from Informal Communication
Focus Areas: Named Entity Extraction, Co-reference Resolution, Document Clustering, Collaborative Filtering
GPA: 4.8/5.0 (includes all classes taken at MIT)
Massachusetts Institute of Technology, Cambridge, MA
M.S. in Electrical Engineering and Computer Science, September 2001
Carnegie Mellon University, Pittsburgh, PA
B.S. in Computer Science, double major in Mathematics, May 1999
Graduated with University and School of Computer Science honors
GPA: 3.9/4.0
Experience
| ActiveState/Sophos, Part-time Consultant | Spam Filtering | Feb 2003 - Sep 2004 |
- Tested machine learning algorithms on internal spam
filtering data sets
- Evaluated competitors' patent applications for soundness
and enforceable inventions
- Taught webinars on modern spam filtering techniques
| ITA Software, Research Scientist | Cache Optimization | Summer 2001 |
- Developed software to analyze and test flight availability cache performance
- Applied machine learning methods to model availability behavior
- Improved caching performance by 170%
| MITRE, Research Scientist | Question Answering | Summer 2000 |
- Developed software for Maximum Entropy statistical model
- Performed candidate ranking within question answering system
- Built infrastructure to support simple training and run-time
prediction for candidate ranking
| MITRE, Research Scientist | Keyword Expansion | Summer 1999 |
- Explored methods for enhancing topic keyword list using WordNet
semantic network
- Conducted information retrieval experiments with keyword
expansion on TREC data using SMART IR engine
- Contributed section of writing to a journal article
| Just Research, Assistant Researcher | Intelligent Web Crawling | Jun 1998 - Dec 1998 |
- Developed web crawler; collected and organized over 20,000 research papers
- Invented methods to efficiently gather topic-specific web documents
- Contributed to design and implementation of research paper search engine
| Carnegie Mellon Univ., Part-time Independent Researcher | Mail Filtering | Jan 1998 - May 1998 |
- Identified need for a program to automatically filter e-mail
based on user preferences
- Developed mail filter program (ifile) utilizing text
classification algorithm
- Released mail filter to public; "ifile" now recognized as first
readily available Bayesian spam filter
- Ran experiments with users to test effectiveness of system
- Analyzed results and explained design of system in research paper
Teaching
| MIT, Teaching Assistant | 6.891 Machine Learning | Sep. 2000 - Dec. 2000 |
- Helped to plan and teach graduate-level Machine Learning course.
- Taught two weekly recitation sessions.
- Contributed to formulation of homeworks and tests.
Professional Activities
- Reviewer for JMLR (various)
- Reviewer for ICML 2007
- Reviewer for NIPS 2005
- Reviewer for SIGIR 2005
- Reviewer for NIPS 2004
- Reviewer for ICML 2004
- Reviewer for SIGIR 2004
- Reviewer for HTL/NAACL 2004
- Reviewer for Spam Conference 2004
- Reviewer for EMNLP 2003
- Reviewer for WWW-2003 Conference
- Reviewer for WWW-2002 Conference
- Reviewer for ICDM-2001 Text Mining Workshop
- Reviewer for KDD-2000 Text Mining Workshop
Academic Honors
- Inducted into Phi Beta Kappa.
- Bachelor's Honor Thesis: "Using Reinforcement Learning to Spider the Web Efficiently"
Other Interests
- Bread
- Wine
- Disc Golf
- Programming (including: ML, C, C++, Python, Perl, PHP, Java, Matlab, Maple, SQL)
Peer-Reviewed Publications
- Loss Functions for Preference Levels:
Regression with Discrete Ordered Labels. [PDF]
Jason D. M. Rennie and Nati Srebro.
Proceedings of the IJCAI Multidisciplinary Workshop on Advances in
Preference Handling. 2005.
- Fast Maximum Margin Matrix Factorization for Collaborative Prediction. [PDF]
Jason D. M. Rennie and Nati Srebro.
Proceedings of the 22nd
International Conference on Machine Learning (ICML). 2005.
- Using Term Informativeness for Named Entity Detection. [PDF]
Jason D. M. Rennie and Tommi Jaakkola.
Proceedings of the 28th
Annual Conference on Research and Development in Information Retrieval
(SIGIR). 2005.
- Maximum-Margin Matrix
Factorization. [PDF]
Nati Srebro, Jason D. M. Rennie
and Tommi
Jaakkola.
Advances in Neural Information Processing Systems (NIPS) 17.
2005.
- TopCat: Data mining for
topic identification in a text corpus. [PDF]
Chris Clifton, Robert
Cooley and Jason Rennie.
Transactions on Knowledge and
Data Engineering 16(8), IEEE Computer Society Press, August, 2004.
- Tackling the Poor Assumptions of
Naive Bayes Text Classifiers. [PDF]
Jason D. M. Rennie,
Lawrence Shih, Jaime Teevan and David R. Karger.
Proceedings of the Twentieth International Conference on
Machine Learning (ICML). 2003.
- Text Bundling:
Statistics-Based Data Reduction. [PDF]
Lawrence Shih,
Jason D. M. Rennie, Yu-Han Chang and David R. Karger.
Proceedings of the Twentieth International Conference on
Machine Learning (ICML). 2003.
- Not Too Hot, Not Too Cold:
The Bundled-SVM is Just Right! [PDF]
Lawrence Shih,
Yu-Han Chang, Jason D. M. Rennie and David Karger.
Proceedings of the ICML-2002
Workshop on Text Learning. 2002.
- Automating the Construction of
Internet Portals with Machine Learning. [PDF]
Andrew McCallum, Kamal
Nigam, Jason Rennie and Kristie Seymore.
Information
Retrieval Journal. Volume 3, pgs. 127-163. Kluwer. 2000.
- ifile: An Application of Machine
Learning to E-mail Filtering. [PDF]
Jason D. M. Rennie.
Proceedings of the KDD-2000
Workshop on Text Mining. 2000.
- Using Reinforcement Learning to
Spider the Web Efficiently. [PDF]
Jason Rennie and Andrew
McCallum.
Proceedings of the Sixteenth International
Conference on Machine Learning (ICML). 1999.
- A Machine Learning Approach to
Building Domain-Specific Search Engines. [PDF]
Andrew McCallum, Kamal
Nigam, Jason Rennie and Kristie Seymore.
Proceedings of
the Sixteenth International Joint Conference on Artificial
Intelligence (IJCAI). 1999.
- Building Domain-Specific
Search Engines with Machine Learning Techniques. [PDF]
Andrew McCallum,
Kamal Nigam, Jason Rennie and Kristie Seymore.
Proceedings of the AAAI Spring Symposium. 1999.
Non-Peer-Reviewed Publications