Movatterモバイル変換

William W. Cohen

Visiting Professor,Bio |Announcements and FAQs |Teaching |Publications (recent,all) |Software |Datasets |Talks |Students & Colleagues |Other Stuff]

Prospective visitors/students: seeannouncements

Biography

William Cohen is a Visiting Professor at Carnegie Mellon University intheMachine Learning Department.He also holds a position as a Principal Scientist at Google, where heworked full-time between May 2018 and March 2024. He received hisbachelor's degree in Computer Science fromDuke University in 1984, and a PhDin Computer Science fromRutgersUniversity in 1990. From 1990 to 2000 Dr. Cohen worked atAT&TBell Labs andlaterAT&T Labs-Research,and from April 2000 to May 2002 Dr. Cohen workedatWhizbang Labs, a companyspecializing in extracting information from the web. From 2002 to2018, Dr. Cohen worked at Carnegie Mellon University intheMachine Learning Department,with a joint appointment intheLanguage TechnologyInstitute.

Dr. Cohen is a past president oftheInternational MachineLearning Society. In the past he has also served as an actioneditor for thetheAIand Machine Learning series of books publishedbyMorgan Claypool, forthejournalMachineLearning, thejournalArtificialIntelligence, theJournal ofMachine Learning Research, andtheJournal of ArtificialIntelligence Research. He was General Chair forthe2008 InternationalMachine Learning Conference, held July 6-9 attheUniversity ofHelsinki,inFinland;Program Co-Chair ofthe2006International Machine Learning Conference; and Co-Chair ofthe1994International Machine Learning Conference. Dr. Cohen was also theco-Chair for the3rdInt'l AAAI Conference on Weblogs and Social Media, which was heldMay 17-20, 2009 in San Jose, and was the co-Program Chair forthe4rd Int'l AAAIConference on Weblogs and Social Media. He isaAAAIFellow, and was a winner of the 2008theSIGMOD"Test of Time" Award for the most influential SIGMOD paper of1998, the2014 SIGIR"Test of Time" Award for the most influential SIGIR paper of2002-2004, and the 2023 Semantic Web ScienceAssociation'sTen-YearAward for the most influential paper of the ISWC-2013 conference.

Dr. Cohen's research interests include include question answering,machine learning for NLP tasks, and neuro-symbolic reasoning, and hehas a long-standing interest in statistical relational learning. Heholds seven patents related to learning, discovery, informationretrieval, and data integration, and is the author of more than 300publications.

Announcements and FAQs

May 2024: A new edition ofA Computer Scientist's Guide To Biology will be out later this summer! More information and an excerpt is available from myco-author's website. I'm possibly biased but I thinkCharles Cohen did a great job with the update - the book is still quite compact, but pretty much the whole book has been rewritten and updated. For example the new version includes several new chapters on topics like CRISPR which weren't even a thing back in 2007.
- On a related note, here's anice non-technical description of LLM hallucinations written by Charlie(based in part on an interview with Vidhisha Balachandran).
March 2024: As you can see from my updated bio above, I am have returned to CMU's ML department full-time (although I still have a 20% involvement a Google, so that email will work!) I'm really looking forward to re-engaging with my friends at colleagues at CMU.
Nov 2023: I'm honored to report that the paperKnowledge Graph Identification, written by Jay Pujara, Hui Miao, Lise Getoor and myself, won a10 year best paper award at the International Semantic Web Conference, 2023.
May 2023: I'm very honored to report that one of the papers I co-authored at EACL 2023 (with Julian Eisenschlos, Jeremy Cole, and Fangyu Liu) won an Outstanding Paper Award.

Teaching

For nowmy old course notes and lectures are avilable through CMU.

Software and demos

Enron email dataset(400Mb, once you get there) contains 800,000+ emails from 150 users+organized into 4700+ folders.

classify.tar.gz (0.4Mb) containsnine problems in which the goal is to classify short entity names.This data was used inJoins that Generalize: Text ClassificationUsing WHIRL (KDD-98).

match.tar.gz (0.7Mb) contains a suite oflabeled entity-name matching and clustering problems(i.e. problems for which the correct matches/clusters are provided),in a single consistent format. In most cases WHIRL's performance isgiven as a benchmark. (These are also distributed in theRIDDLERepository. Extraction-oriented versions of some of this data areavailable on theRISERepository. (I.e., represented as a problem of extracting data froma website, rather than matching two datasets).)

whirl-bench.tgz (1.1Mb) contains somemore WHIRL-format entity name matching problems.

Talks and presentations

Video for an invited talk, KR for KBQA and KBC given at AKBC, June 2020.(This is the whole first day of the conference, my talk starts about 17:42, after the opening remarks.)
An invited talk given attheThe 16th InternationalConference on Principles of Knowledge Representation on Oct 27th -Nov 2nd 2018, in Tempe, Arizona.
An invited talk given atthe28th International Conferenceon Inductive Logic Programming on September 2nd - 4th 2018, inFerrara, Italy (there'salsovideofor this talk).
An invited talkgiven at Third International Workshop on Declarative Learning BasedProgramming (DeLBP), at AAAI-2018.
An invited talk given at SNL-2017 (the 1st International Workshop on Symbolic-Neural Learning) in July 2017.
An invited talk given at WAKBC-2016 in June 2016.
Tutorial on statistical relational learning given at NAACL 2016 withWilliam Wang (a shorter version of this was also presented at IJCAI 2016):
- Part 1 - overview on logic, probability, MLNs, and probabilistic DDBs
- Part 2 - ProPPR and applications
- Part 3 - TensorLog, and other recent and current work
Series of three lectures on probabilistic logic programs given atSingapore Management University in Feb 2016:
- Background on logic and probabilistic models
- Parameter learning and structure learning in ProPPR
- Joint learning in ProPPR and comparing to neural approaches
Can KR Represent Real-World Knowledge?, invited talk given March 2015at the AAAI Spring Symposium on Knowledge Representation and Reasoning: Integrating Symbolic and Neural Approaches
Learning to Reason with Extracted Information, keynote talk given March 2014at Google's Natural Language Understanding Workshop, Zurich, Switzerland.
Learning to Construct and Reason with aLarge KB of Extracted Information, invited talk given August 2013at the Inductive Logic Programming Conference, in Rio de Janeiro,Brazil.
Reasoning With Data Extracted from The Biomedical Literature,invited talk at a joint session of the AAAI Fall Symposia on Discovery Informatics, andInformation Retrieval and Knowledge Discovery in Biomedical Text.
Learning Similarity Relations Based on Random Walks in Graphs,invited talk at CIKM 2012, October, 2012.
- Earlier version of talk:Learning Relationships Defined byLinear Combinations of Constrained Random Walks, invited talk at the9th Workshop onMachine Learning and Graphs, San Diego, CA, Aug 2011.
Fast Effective Clustering for Graphs and Documents, given at CMU's LTI Colloquium Feb 10, 2012.
- Earlier versions given atVirginia Tech in April 2010 andUniversity of Pennsylvania in Feb 2010.
Learning to Extract a Broad-CoverageKnowledge Base from the Web, invited talk at the Symposium onData-Intensive Analysis, Analytics, and Informatics, Pittsburgh, PA Apr 2011.
Open Information Extraction Methods:Computers that Learn to Read, invited talk at National Federationof Advanced Information Services (NFAIS), Philadelpha, PA, Feb 2011.
Learning Proximity Relations Defined byLinear Combinations of Constrained Random Walks, given at aseminar at the University of Maryland in Sep 2010.
Modeling Entity-Entity Linksand Entity-Annotated Text, given at the ICML 2010 Workshop onTopic Modeling.
Predictively Modeling Social Media,invited talk given atthe 1st International Workshop on Mining Social Media, co-located with 13th Conference of the Spanish Association for Artificial Intelligence (CAEPIA-TTIA 2009).
Matching and clustering product descriptionsusing learned similarity metrics, invited talk given atthe IJCAI 2009 Workshop on Information Integration on the Web, July 2009. (Powerpoint; 6.7M)
Open information extraction talks:
- Graph-Based Methods for Open Information Extraction, talk given at Nov 2009 at MAVIR in Madrid, Spain.
- Graph-Based Methods for Open Information Extraction, earlier version of talk given at Stanford and Google March 2009.
- Graph-Based Methods for Open Information Extraction, still earlier version of the same talk given at a 2008 NIPS workshop.
- AQT video of highlights from the workshop talks, including an incisive technical question addressed to me from my colleagueSteve Fienberg.
Embodied Cognition and Knowledge:Integration of Heterogeneous Databases without Common Domains UsingQueries Based on Textual Similarity, talk given for my 10-year"Test of Time" Award atSIGMOD-2008(Powerpoint; 11Mb)
Using Machine Learning to Discoverand Understand Structured Data, invited talk given atLinkedData2008. (Powerpoint; 6Mb)
Machine Learning for Personal InformationManagement, invited talk given atICMLA-2007. (Powerpoint; 8Mb)
A Framework for Learning to Query Heterogeneous Data,invited talk given atIQIS 2006. (Powerpoint; 8Mb)
On Beyond Hypertext: Searching in GraphsContaining Documents, Words, and Actual Data, invited talk givenatDB/IR Day 2006. (Powerpoint; 6Mb)
A Century Of Progress On InformationIntegration: A Mid-Term Report, an overview of informationintegration, focusing modestly on my own work, given as invitedtalk atWebDB-2005. (Powerpoint;12Mb)

Tutorials:
- Information extraction (PowerPoint; 4.8Mb), aimed at folks somewhat familiar with statistical NLP methods. And thanks to Thierry Poibeau, there's also a versionen francais (did I get that right, Thierry?) Also, two earlier versions of this are also still around, both given with Andew McCallum at recent conferences,KDD-2003(PowerPoint; 6.8Mb) andNIPS-2002.
- Text classification (PowerPoint; 3Mb), given at a CALD Summer Course.
- Collaborative filtering (PowerPoint; 9.1Mb), given at a DIMACS workshop.
A mini-course on record linkage and matching:
- Overview of record linkage methods(PowerPoint; 250kb).
- Overview of distance metrics for strings(PowerPoint; 530kb).
- Overview of using HMMs for normalizingtext in record linkage tasks(PowerPoint; 640kb).
  It's not a presentation, but I have also put together ashort annotated bibliography of record linkage and matching papers.
- William Hayes has a nice summary ofan extended discussion of string-matching tools on the BioNLP mailing list (July 2012).
Other technical talks:
- A presentation of my IJCAI-2005 resultson "stacked sequential learning", presented in Edinburgh in August, 2005.
- A presentation of my NIPS-2002 resultson using bootstrapping techniques to improve web page classification,given at CMU in October 2002. (PowerPoint; 3.2mb).
- A presentation of my WWW-2002 resultson wrapper learning,presented in April 2002. (PDF; 170kb).
- An overview of experiments with WHIRL. (PDF; 800kb).

Publications

Recent and selected publications.
All publications. Here is an more-or-lesscomplete chronological list of my publications.
Possibly more up-to-date: here are pointers tomyScholar page,andmyArxiv page.
The second editionofAComputer Scientist's Guide To Cell Biology, 2nd Edition is out,with better graphics and many updates. (If you are an academic you canprobably get the ebook free from Springer.) The second edition isstill a compact guide to biology written from the perspective ofcomputer science, but includes new updated material, better graphics,and punchier writing, thanks tomyco-author, a young sciencewriter who worked for many years doing hands-on wetlab biology.
Students and other colleagues
Current students:
- Daniel Spokoyny, LTI PhD student, co-supervised with Taylor Berg-Kirkpatrick.
Former students/colleagues:
- Katie Rivard Mazaitis, research programmer/analyst
- Jay Yoon Lee, CSD PhD student, co-advised with Jaime Carbonell, now a postdoc at UMass/Amherst.
- Bhuwan Dhingra, LTI PhD student, co-advised with Ruslan Salakhutdinov, now an Assistant Professor at Duke University.
- Zhilin Yang, former LTI PhD student, co-advised with Ruslan Salakhutdinov, now an Assistant Professor at Tsinghua University.
- Rose Catherine Kanjirathinkal, former LTI PhD student, now at Yahoo.
- Fan Yang (former MLD PhD student)
- William Yang Wang, former LTI PhD student, now an Assistant Professor at UCSB.
- Dana Movshovitz-Attias former CSD PhD student, now at Google.
- Bhavana Dalvi Mishraformer LTI PhD student(co-advised withJamie Callan, now at AI2.
- Tae Yano, former LTIPhD student, co-advisedwithNoah Smith.
- Nan Li, former CSD PhDstudent, co-advisedwithKenKoedinger, now at D. E. Shaw.
- Ramnath Balasubramanyan, former LTI PhD student.
- Mahesh Joshi, former LTI PhD student,co-advised withCarolyn Rosé
- Frank Lin, former LTI PhD student
- Ni Lao former LTI PhD student
- Richard C. Wang, former LTI PhD student co-advised withBob Frederking.
- Andrew Arnold former MLD PhD student, now at Amazon.
- Einat Minkov former LTI PhD student, now at Haifa University.
- Vitor Rocha de Carvalho former LTI PhD student, now at SnapChat.
- Zhenzhen Kou former MLD PhD student, now at Google.
- Yifeng Tao, CMU Comp Bio PhD student, now supervised by Russell Schwartz.
- Edoardo Airoldiformer MLD/Stats PhD student, co-advised withSteve Fienberg, now at Harvard.
- Pradeep Ravikumarformer MLD PhD student, co-advised withSteve Fienberg, now at CMU.
- Haitian Sun, former MLD MS student, now a PhD student at CMU.
- Fan Yang, MLD MS student.
- Qiao Jin, School of Medicine, Tsinghua University, now supervised by Xinghua Lu.
- Ezra Winston, former MLD Master's student, now a PhD student in MLD.
- Lanxio (Karen) Xu, former MLD Master's student.
- Yuxing Zhang, former MLD Master's student.
- Jakob Bauer, former MLD 5th-year Master's student, now at Google..
- Kavya Srinet, former MCDS Master's student.
- Bhawna Juneja, former MCDS Master's student.
- Tom Shen, former CMU CSD undergrad.
- Yu-Hsin Allen Kuo, former LTI MLT student, formerly co-advised withNatasa Miskov-Zivanov
- Rahul Goutam, former LTI MLT student, co-advised withNatasa Miskov-Zivanov
- Malcolm Greaves, former CSD master's student.
- Wen Haw Chong, PhD student at Singapore Management University,visted CMU in 2015-2016.
- TuanAhn Hoang, PhD student at Singapore Management University,visited CMU for 2012-2013 academic year.
- FreddyChong Tat Chua PhD student at Singapore Management University,visited CMU for the academic year 2011-2012.
- Gustavo Lacerdaformer research assistant, co-supervised with Noboru Matsuda and Ken Koedinger.
- Lidong Bing, formerpostdoc.
- Ramesh Nallapatiformer postdoc, co-supervised withJohn Lafferty.
- Noboru Matsudaformer postdoc, co-supervised withKen Koedinger.
- Ja-Hui Changvisiting faculty from National Central University, Taiwan, 2007-2008.
- I have been an external committee member for the PhD theses of
  - John Zelle (degreefrom U Texas)
  - MishaBilenko (from U Texas)
  - Daniel Kudenko(Rutgers)
  - Chumki Basu (Rutgers)
  - Ananlada Chotimongkol (CMU)
  - Wei-Hao Lin (CMU)
  - Cenk Gazen (CMU)
  - David Nadeau (U Ottowa)
  - Hanghang Tong (CMU)
  - Ben van Durme (Rochester)
  - Partha Talukdar (U Penn)
  - Andy Carlson (CMU)
  - Yifen Huang (CMU)
  - Swapna Sundaran (UPitt)
  - Michael Heilman (CMU)
  - Jon Elsas (CMU)
  - Dipanjan Das (CMU)
  - Fan Guo (CMU)
  - Jana Diesner (CMU)
  - Freddy Chong Tat Chua (Singapore Management University).
  - Qirong Ho (CMU)
  - Danai Koutra (CMU)
  - Reyyan Yeniterzi (CMU)
  - YiChi Wang (CMU)
  - Steven Gardiner (CMU)
  - Jay Pujara (Univ Maryland)
  - Derry Wijaya (CMU)
  - Lingjia Deng (Univ of Pittsburgh)
  - Chenyan Xiong (CMU)
  - Tiancheng Zhao (CMU)
  - Pradeep Dasigi (CMU)
  - Shashank Srivastava (CMU)
  - Abulhair Saparov (CMU)
  - Danish Pruthi (CMU)
  I have also been an external committee member for the Master's theses ofMehrbod Sharifi (CMU) andWeam Abu-Zaki (CMU).
  I am currently a PhD committee member for Vidhisha Balachandran,Zhengbo Jiang, Luyu Gao, and Sankey Vaibhav Mehta.
  Other Stuff
  For those many friends whose research I have built on, be warned.My full name, "William Weston Cohen", is an anagram of the phrase "Inow cite shallow men". (FromSara Cohen - norelation! - comes this warning: "Women's rights activists wouldprobably request you to use the following anagram instead: 'I shallnow cite women'".)
  Through my advisor, Alex Borgida, I can tracemy"academic lineage" back to luminarieslike Leibniz, Newton and Alfred Whitehead.Update: My formerstudentAndrew Arnold has gone backevenfurther, toGalileo!
  In 2014 I unearthed a strange relic from the past, a sort ofgame/website I wrote for mysonCharlie back in...I'mgonna say, 1994, 1995, something like that, and I sort of made it workagain, although JavaScript has changed a bit in the last couple ofdecades. (The main bugs have to do with sound-file presentation - in1994 these were played by mime-file configured helper programs, notnatively by the browser, so now you need to hit 'back' about 1/2 thetime after a sound plays.)Historicallyinteresting? You decide!
  Poetry anyone?
  
  [8]ページ先頭
  
  ©2009-2025 Movatter.jp