Gregory Piatetsky-Shapiro Chair, SIGKDD gregory@KDnuggets.com

Document Sample
Gregory Piatetsky-Shapiro Chair, SIGKDD gregory@KDnuggets.com Powered By Docstoc
					     Gregory Piatetsky-Shapiro
          Chair, SIGKDD
     gregory@KDnuggets.com

   ACM Special Interest Group on
Knowledge Discovery and Data Mining
         www.SIGKDD.org



           KDD-09 Business Meeting, Jun 30, 2009
                 What is Data Mining?
• Making sense of Data

• The Art and Science of finding useful patterns in data

• The non-trivial process of identifying valid, novel,
  potentially useful and ultimately understandable
  patterns in data (from 1996 book by Fayyad,
  Piatetsky-Shapiro, Smyth).


                    KDD-09 Business Meeting, Jun 30, 2009
               SIGKDD – what does it do?
• The Professional Society for Data Mining and
  Knowledge Discovery (ACM SIG)

• Organize the annual KDD Conference
   – Co-sponsor WSDM - Web Search and Data Mining
   – Support (in-cooperation) with other conf.
      • ICUIMC, SPB, RECSYS, …


• Publish SIGKDD Explorations


                      KDD-09 Business Meeting, Jun 30, 2009
                              Awards
• Annual Awards:
   – Innovation – the “Nobel” prize in Data Mining
   – Service: Extraordinary Service to the community
   – Outstanding Dissertation

• Other activities
   – Data Mining Curriculum published
   – Webcasts
   – KDD-09 Social network – kdd09.crowdvine.com



                     KDD-09 Business Meeting, Jun 30, 2009
                         Our Goal:
– Facilitate Research in Data Mining
– Improve Industry – Academia Interaction
– Support Data Mining Community




                  KDD-09 Business Meeting, Jun 30, 2009
                     Agenda
•   SIGKDD progress and Initiatives
•   KDD – 20 years after
•   TKDD report (Jiawei Han)
•   SIGKDD Explorations (Osmar Zaiane)
•   New SIGKDD Chair (Usama Fayyad)
•   KDD-10 in Washington, D.C. (?)


                 KDD-09 Business Meeting, Jun 30, 2009
            Communicating KDD News

Visit www.KDnuggets.com

Subscribe to KDnuggets News (twice/month)

Follow                  .com/kdnuggets


For Data Mining and Knowledge Discovery news

                 KDD-09 Business Meeting, Jun 30, 2009
              20 Years After (Vingt ans après)
• 1989: KDD-89 – First Workshop on Knowledge
  Discovery and Data Mining, Detroit (KDD Stone age)
Important research directions in 1989:
   –   Domain knowledge.
   –   Dealing with Uncertainty
   –   Efficiency.
   –   Incremental Approaches.
   –   Interactive Systems.
• Large Database in 1989: 1 MB
• Median Largest DB in 2009: 10 TB (KDnuggets poll)
                         KDD-09 Business Meeting, Jun 30, 2009
                   KDD Bronze age
• 1991, 93, 94 Workshops

• 1995: KDD-95 – First
  Conference on Knowledge
  Discovery and Data Mining,
  Montreal




                   KDD-09 Business Meeting, Jun 30, 2009
                           SIGKDD
• 1999: SIGKDD created

• Won Kim, Chair (1999-2005)

• Gregory Piatetsky-Shapiro, Chair (2005-2009)




                  KDD-09 Business Meeting, Jun 30, 2009
                 KDD Today – modern age
• 25 Data Mining / KDD conferences in 2009
   – KDD, mainly US (sometimes Europe, ?)
   – IEEE ICDM: rotates US/Asia/Europe
   – SIAM SDM – US
• Regional
   – ECML-PKDD - Europe
   – PAKDD – Pacific Asia
• National: Australia, China, France
   – Vendor conferences: SAS, SPSS, Salford, …
   – Commercial conferences: Predictive Analytics World
• Many thousands of papers
                        KDD-09 Business Meeting, Jun 30, 2009
                  New Research Areas
•   Social Networks
•   Text Mining
•   Mobile / geographic / sensor data
•   Privacy
•   Image/multi-media mining
•   …

• Many deployed applications - embedded

                    KDD-09 Business Meeting, Jun 30, 2009
Term                      Count
                                                        Top 2-word terms in
data mining                                  18
                                                         KDD-09 abstracts
experimental result                          15
                                                            - thanks to
social network                               13
learning algorithm                             7
                                                            Gabor Melli
machine learning                               7
recommender system                             7
search engine                                  7
labeled data                                   6
learning method                                6
training data                                  6
collaborative filtering                        5
                          KDD-09 Business Meeting, Jun 30, 2009
Analytics Surpassing Data Mining ?
          (Google Trends)




 www.kdnuggets.com/news/2009/n02/3i.html

             KDD-09 Business Meeting, Jun 30, 2009
                  SIGKDD Membership
• Total membership as of June’09 is 1,161

• Membership overall steady, slight decline
   – student membership slightly up in 2009




                     KDD-09 Business Meeting, Jun 30, 2009
SIGKDD vs SIGMOD vs SIGIR membership




   All SIGS – 7% decline from 2000 to 2008
              KDD-09 Business Meeting, Jun 30, 2009
SIGKDD vs SIGMOD vs SIGIR Student members




                                                         SIGKDD
                                                         Student
                                                         members
                                                         slightly up in
                                                         2009




      All SIG student members – 43% decline from 2000 to 2008
                 KDD-09 Business Meeting, Jun 30, 2009
                            Instant Poll
• Are you a SIGKDD Member?
• Benefits
   –   Big discount on KDD registration
   –   Explorations hardcopy
   –   Vote in next election
   –   Belong to the largest KDD professional organization
• Cost: $25 ($15 / students)
• Join at www.SIGKDD.org



                       KDD-09 Business Meeting, Jun 30, 2009
                        Instant Poll
• Researcher (University, Lab)

• Student

• Industry/Government/Consultant




                   KDD-09 Business Meeting, Jun 30, 2009
               KDD-09 in Paris
• Why Paris?


• Paris vaut bien une messe
  –Henri IV
  (Paris is worth it, despite a lot of messy
    organization problems)

               KDD-09 Business Meeting, Jun 30, 2009
Conference Attendance


                                           KDD-09
                                           764 as
                                           of Jun 29




   KDD-09 Business Meeting, Jun 30, 2009
USA
France
                 271
                 112
                                                 By Country
Germany           58
China             36
Belgium           27
United Kingdom    27
Canada            18
Japan             16
Spain             14
Netherlands       13
Slovenia          13
Israel            11
Finland            9
Italy              9
Singapore          9
Switzerland         9
Australia           8
Brazil              8
Austria             7
Taiwan              7
                 KDD-09 Business Meeting, Jun 30, 2009
         Cash Balance of SIGKDD

$1 M as of June 30, 2009




             KDD-09 Business Meeting, Jun 30, 2009
        Crowdvine – KDD-09 Social network
• Feedback ?

• Many thanks to
  – Gabor Melli (PredictionWorks)
  – Sankar Virdhagriswaran (The Hartford)
  – Kunal Punera (Yahoo)




                    KDD-09 Business Meeting, Jun 30, 2009
WSDM-2010 Call-for-Paper




           25
                    WSDM2010.org
• WSDM is a young ACM conference focused on Web search
  and data mining. It has been sponsored by SIGKDD, SIGIR,
  and SIGWEB.
   – Submission deadline: Aug 13, 2009
   – Notification: Oct 15, 2009
• Hope you will submit your papers!




                             26
Report on ACM Transactions on
Knowledge Discovery from Data

                  Jiawei Han
               Editor-in-Chief
University of Illinois at Urbana-Champaign

             KDD-09 Business Meeting, Jun 30, 2009
  SIGKDD
Explorations


  KDD-09 Business Meeting, Jun 30, 2009
 ACM SIGKDD Explorations
Official newsletter of the SIGKDD
community.

 http://www.kdd.org/explorations


Editor-In-Chief :

                Osmar R. Zaïane


Associate Editors:

                Charu Aggarwal



                Bart Goethals
                                 KDD-09 Business Meeting, Jun 30, 2009
                  About SIGKDD Explorations
• We have two issues a year (typically June & December).
• All issues are available online
  http://www.kdd.org/explorations/
• Contains
    –   Submitted articles (reviewed)
    –   Reports (workshops, kdd cup, and others)
    –   Interviews
    –   Selected KDD workshop papers
    –   Book Reviews
    –   Doctoral Dissertation Abstracts
    –   Special issues guest edited
• Members get hard copy.

                           KDD-09 Business Meeting, Jun 30, 2009
 Upcoming Issues:                   Special Issues
• Educational Data Mining (in negotiations) 06/10
• Visual Analytics &KD [Kai Puolamäki] 12/09
 Past Issues:
• Open Source Data Mining Software                            [R. Grossman] 07/09
•   Visual Analytics                                         [D. Keim, J. Schneidewind] 12/07
•   Data Mining for Health Informatics                       [Raymond T. Ng, Jian Pei ] 06/07
•   Successful Real-World Data Mining Applications [G. Melli, O. R. Zaiane, B. Kitts ] 06/06
•   Link Mining                                              [L. Getoor and C. P. Diehl ] 12/05
•   Natural Language Processing and Text Mining              [Anne Kao, Steve Poteet ] 06/05
•   Web Content Mining                                       [B. Liu, K. C.-C. Chang ] 12/04
•   Learning from Imbalanced Datasets            [N. Chawla, N. Japkowicz, A. Kolcz ] 06/04
•   Microarray Data Mining                       [G. Piatetsky-Shapiro, P. Tamayo ] 12/03
•   Multi Relational Data Mining                             [S. Dzeroski, L. De Raedt ] 06/03
•   Privacy and Security                                     [J. Gehrke] 12/02
•   Constraints in Data Mining                               [R. Bayardo] 06/02
•   Online, Interactive, and Anytime Data Mining             [M. Zaki] 01/02
•   Scalable Data Mining                                     [K. Shim] 12/00
•   Internet Data Mining                                     [P. Bradley] 06/00
              We invite…                                                Send to:
                                                                editors_sigkdd@acm.org

• Reports on competitions, panels, workshops:
    – Deadline: Oct 23rd, 2009.
•   Proposals for special issues.
•   Articles with industry and application focus.
•   Technical articles on the special issue
•   Position papers on controversial issues
•   Commentaries/notes requiring quick dissemination
•   Book reviews, Advertisements
•   PhD dissertation abstracts (2 pages)
                        KDD-09 Business Meeting, Jun 30, 2009
              KDD-09 Transfer Meeting
• Tonight 7 pm, La Seine.




                   KDD-09 Business Meeting, Jun 30, 2009
                     SIGKDD Outgoing
                         Officers
                        2005-2009
Gregory Piatetsky-Shapiro,                                 Usama Fayyad,
Chair                                                      Secretary/Treasurer
                              Directors




                  David             Jiawei                Sunita
 Robert                                                                  Raghu
                  Jensen            Han                   Sarawagi
 Grossman                                                                Ramakrishnan
                             KDD-09 Business Meeting, Jun 30, 2009
                       SIGKDD Newly
                      Elected Officers

Usama Fayyad, Chair                                       Osmar Zaiane,
                                                          Secretary/Treasurer
                         Directors




Johannes   Robert       David                Sunita             Raghu        R.
Gehrke     Grossman     Jensen               Sarawagi           Ramakrishnan
                                                                             Srikant
                        KDD-09 Business Meeting, Jun 30, 2009