OSDM'09: The first Open Source in Data Mining workshop

Bangkok, Thailand, 27 April 2009  

Time and venue: The Imperial Queen's Park Hotel, Bangkok, Room #3, 13:00 - 18:30.

Registration: Please use the PAKDD'09 Registration System.

Open source software is becoming increasingly accepted in public and private sector organisations in many countries. There is a variety of open source data mining tools available to both researchers and practitioners, some being simple research prototypes while others are fully developed software tools.

This workshop aims to bring together data mining practitioners, researchers and educator, with the objectives to present open source data mining tools, discuss experiences and lessons learned using such tools, and exchange ideas on how to promote the use of open source tools in the field of data mining.

Workshop Program

1. 13:00 - 13:10: Opening and Welcome.

2. 13:10-14:10: Invited keynote presentation by Dr Mark Hall, Pentaho, New Zealand, one of the original core developers of the WEKA open source data mining system.

   Title: The WEKA open source data mining system

   Abstract:   More than twelve years have elapsed since the first public release of WEKA. In that time, the software has been re-written entirely from scratch, evolved substantially and now accompanies a text on data mining. These days, WEKA enjoys wide-spread acceptance in both academia and business, has an active community, and has been downloaded more than 1.4 million times since being placed on SourceForge in April 2000. In this talk, I'll review the history of WEKA, and, in light of the recent 3.6.0 stable release, briefly discuss what has been added since the last stable version in 2004.

   Biography:   Mark Hall is one of the original core developers of the WEKA data mining software and is responsible for leading Pentaho's data mining solutions. He has 15 years experience as an academic researcher in computer science and has published extensively in machine learning and data mining conferences and journals. Prior to joining Pentaho, Mark held teaching and postdoctoral fellowship positions at the University of Waikato in New Zealand.

3. 14:10-15:10: Technical papers (part 1)
   1. OpenSubspace: An Open Source Framework for Evaluation and Exploration of Subspace Clustering Algorithms in WEKA.
      Emmanuel Mueller, Ira Assent, Stephan Guennemann, Timm Jansen, Thomas Seidl.
      Software is available online.

   2. The open source library iZi for pattern mining problems.
      Frederic Flouvat, Fabien De Marchi, Jean-Marc Petit.

4. 15:10-15:30: Afternoon tea/coffee break

5. 15:30-16:30: Technical papers (part 2)
   3. The Konstanz Information Miner 2.0.
      Thorsten Meinl, Nicolas Cebron, Thomas Gabriel, Kilian Thiel, Bernd Wiswedel, Michael Berthold, Fabian Dill, Peter Ohl, Tobias Koetter.

   4. Cougar2: An Open Source Machine Learning and Data Mining Development Platform.
      Abraham Bagherjeiran, Oner Celepcikay, Rachsuda Jiamthapthaksin, Chunsheng Chen, Vadeerat Rinsurongkawong, Seungchan Lee, Justin Thomas, Christoph Eick.

6. 16:30-18:00: Demonstration session of open source data mining tools

   The following tools will be demonstrated (30 minutes each):

   1. KNIME
   2. WEKA
   3. Rattle

7. 18:00-18:30: Panel discussion of open source developers

   Topic will be one (or both) of:

   • Why open source for data mining research and education?
   • Why publish your data mining tool as open source software?

   This will be an informal session where workshop participants are encouraged to participate.

   Confirmed panel participants:

   • Dr Mark Hall, Pentaho, New Zealand (WEKA developer)
   • Dr Graham Williams, Togaware, Australia (Rattle developer)
   • Dr Nicolas Cebron, University of Konstanz, Germany (KNIME developer)

Important Dates

Submission of full papers:	9 January 2009     (passed)
Notification of Authors:	30 January 2009     (passed)
Camera-ready version:	16 February 2009     (passed)
OSDM'09 workshop date:	27 April 2009, 13:00-18:30

• Dr Peter Christen, The Australian National University.
• Dr Graham Williams, Togaware, Australia.

• Dr Rohan Baxter, The Australian Taxation Office, Australia
• Prof Michael Berthold, University of Konstanz, Germany
• Dr Christian Borgelt, European Center for Soft Computing, Spain
• Assist Prof Janez Demsar, University of Ljubljana, Slovenia
• Assoc Prof Eibe Frank, University of Waikato, New Zealand
• Dr Mark Hall, Pentaho, New Zealand
• Prof Joshua Huang, The University of Hong Kong, Hong Kong
• Assoc Prof Bernhard Pfahringer, University of Waikato, New Zealand
• Assoc Prof Blaz Zupan, University of Ljubljana, Slovenia
• Dr Yunming Ye, Harbin Institute of Technology, China

• KNIME: a modular data exploration and mining platform.
• Orange: A Python based data mining toolkit.
• Rattle: A data mining GUI written in the statistical software environment R and available as a plugin for the WebFocus BI product.
• WEKA: A comprehensive data mining system written in Java and sponsored by the open-source BI software company Pentaho.
• OpenSubspace: A subspace clustering extension to WEKA.
• Febrl (Freely Extensible Biomedical Record Linkage): A data cleaning, deduplication and data linkage system.