ICDM'2012
Comments from Participants of ICDM
Comment from Dr. Graham Williams
A Data Mining Showcase: An Overview of ICDM 2012 at DSA 2012, Berlin
Data Mining continues its growth in supporting Industry, Government, and Research. Using a data driven approach to building models based on massive data collections, data mining models are becoming universally deployed.
The 12th Industrial Conference on Data Mining ICDM 2012, held in Berlin, July 2012, showcased many innovative and challenging applications of data mining. The conference was co-located with and incorporated within the World Congress on Data and Signal Analysis, DSA 2012. It was held 16 July - 20 July in the spectacular and historic city of Berlin, attracting some 110 attendees from over 40 countries.
In addition to ICDM, DSA 2012 also hosted the 7th International Conference on Mass Data Analysis of Images and Signals (MDA 2012), covering applications in medicine, biotechnology, food industries, dietetics, biometry, security and agriculture. The 8th International Conference on Machine Learning and Data Mining (MLDM 2012) rounded out the agenda for DSA 2012.
The conference program allowed attendees to participate in each of the co-located conferences through a single stream. This provided a unique opportunity to hear from speakers across a broad spectrum of areas, all related to the analysis of big data. Interaction between the participants from different disciplines and interests ensured lively discussion, questioning, and sharing of insights and experiences.
Maintaining a high standard with an acceptance rate of around just 30%, the quality of the research reported at the conference remains very high. The proceedings are published by Springer as Advances in Data Mining. Applications and Theoretical Aspects - 12th Industrial Conference, ICDM 2012, LNCS, vol. 7377 ISBN 978-3-642-31487-2. It is edited by Professor Petra Perner.
The poster and industry proceedings are published by ibai-publishing as Advances in Data Mining, Poster and Industry Proceedings - 12th Industrial Conference, ICDM 2012, ISBN 978-3-942952-17-0. It is edited by Professor Petra Perner.
The workshop proceedings for the Workshops: Workshop on Case-Based Reasoning CBR-MD 2012, Workshop on Data Mining in Agriculture, DMA 2012, Workshop on Data Mining in Life Sciences DMLS 2012, and Workshop on Data Mining in Marketing DMM 2012 are published as Advances in Data Mining, Workshop Proceedings, 12th Industrial Conference, ICDM 2012, ISBN 978-3-942952-16-3. It is edited by Isabelle Bichindaritz, Petra Perner, Georg Ruß and Rainer Schmidt.
A number of invited talks across the conferences focused on practical applications of the technology. Professor Xiaoqing Ding of Tsinghua University, China, fascinated the attendees with a technology update on face recognition and image processing. Their systems and algorithms have been in daily use for many years for automated immigration processing, for example, for the border crossing between Shenzhen and Hong Kong, with near perfect accuracy.
Professor Patrick Wang was another highlight. Having once had an office at MIT between Marvin Minsky on one side and Patrick Winston on the other, Professor Wang was well and truly placed in the centre of Artificial Intelligence research. We were presented with a tour de force of some fascinating practical applications of AI, including image analyses related to understanding Chinese characters.
Professor Petra Perner presented a new view to System Biology during her talk "Quantitative Measurement of Cellular Events". She highlighted the need to consider this problem in a more systemic view. Such an approach will surely help to better understand characteristics the assays should have and the kind of measurements necessary in order to standardize these assays. Progress will also allow these tasks to be classified under a system input and output view, leading to new procedures that can extract the necessary high-level information from the microscopic images. Professor Perner presented examples that follow this view and provide new insights into the application of the approach.
I was delighted to have the opportunity to present an invited talk covering "Aspects of Data Mining" in practise. The talk gave an overview of the important data mining procedures introduced for the application of discovering knowledge from taxation data. The talk showed how much has been achieved so far in order to be able to mine these data and presented new procedures based on building ensembles of decision trees to efficiently support this task.
The conference banquet celebrated many of the most prominent participants of the conference series. A group of 12 of them received a special commemorative award featuring the famous cartographer, and scientific entrepreneur, Gerardus Mercator.
Another common feature of the conference series is the recognition of the top research being presented. It is often difficult to single out specific papers from amongst many excellent pieces of research, but the organisers had identified the best of the best, and highlighted these in the program. The top offerings included a paper by Gabor Nagy and Krisztian Buza on the efficient storage of tick data for search and analysis, a paper by Jef Vanlaer, Pieter Van den Kerkhof, Geert Gins, and Jan F.M. Van Impe on the influence of input and output measurement noise on batch-end quality prediction with partial least squares, and the winning paper by Leon Bobrowski and Tomasz Lukaszuk on prognostic modeling with high dimensional and censored data.
ICDM 2012 was a great success and comes highly recommended as a conference that brings research to industry and presents a showcase of many actual applications of data mining. The next conference will be held 16-21 July 2013 in New York (visit http://www.data-mining-forum.de for details). The track record of the conference series suggests that DSA 2013 will be a high quality and informative conference.
Dr. Graham Williams
Senior Director and Chief Data Miner, Australian Taxation Office
International Expert and Visiting Professor, Chinese Academy of Sciences
Adjunct Professor, Data Mining, Fraud and Security, University of Canberra
Adjunct Professor, Australian National University
