Keynote Speaker

 

Prof. Christos Faloutsos Department of Computer Science, Carnegie Mellon University, U.S.A. Homepage: http://www.cs.cmu.edu/~christos/ Title: Graph Mining: Laws, Generators and Tools   Abstract: How do graphs look like? How do they evolve over time? How can we generate realistic-looking graphs? We review some static and temporal 'laws', and we describe the "Kronecker" graph generator, which naturally matches all of the known properties of real graphs. Moreover, we present tools for discovering anomalies and patterns in two types of graphs, static and time-evolving. For the former, we present the 'CenterPiece' subgraphs (CePS), which expects $q$ query nodes (e.g., suspicious people) and finds the node that is best connected to all $q$ of them (e.g., the master mind of a criminal group). We also show how to compute CenterPiece subgraphs efficiently. For the time evolving graphs, we present tensor-based methods, and apply them on real data, like the DBLP author-paper dataset, where they are able to find natural research communities, and track their evolution.  Finally, we also briefly mention some results on influence and virus propagation on real graphs.   Short Biography: Christos Faloutsos is a Professor at Carnegie Mellon University. He has received the Presidential Young Investigator Award by the National Science Foundation (1989), the Research Contributions Award in ICDM 2006, ten ``best paper'' awards, and several teaching awards. He has served as a member of the executive committee of SIGKDD; he has published over 160 refereed articles, 11 book chapters and one monograph. He holds five patents and he has given over 20 tutorials and 10 invited distinguished lectures. His research interests include data mining for streams and graphs, fractals, database performance, and indexing for multimedia and bio-informatics data.

Invited Speakers

(Sorted in alphabet order)

 

Prof. Hiroki Arimura Division of Computer Science, Hokkaido University, Japan Homepage: http://www-ikn.ist.hokudai.ac.jp/~arim/arim_en.html Title: Efficient Algorithms for Mining Frequent and Closed Patterns from Semi-structured Data   Abstract: In this talk, we study efficient algorithms that find frequent patterns and maximal (or closed) patterns from large collections of semi-structured data. We review basic techniques developed by the authors, called the rightmost expansion and the PPC-extension, respectively, for designing efficient frequent and maximal/closed pattern mining algorithms for large semi-structured data. Then, we discuss their applications to design of polynomial-delay and polynomial-space algorithms for frequent and maximal pattern mining of sets, sequences, trees, and graphs.   Short Biography: Hiroki Arimura is a Professor in the Graduate School of Information Science and Technology of Hokkaido University. After receiving his PhD from Kyushu University in Computer Science, his primary interests are in knowledge discovery and combinatorial pattern matching for semi-structured data, computational learning theory, design and analysis of algorithms, and bioinformatics. He has served on the program committees of a number of international conferences including being a program co-chair for Algorithmic Learning Theory conference in 2000 and a member of the steering committee of Discovery Science conference series since 2006. Since 2007, he has been the director of the Global COE (Centers of Excellence) Program of "Center for Next-Generation Information Technology Based on Knowledge Discovery and Knowledge Federation" at Hokkaido University by MEXT (Ministry of Education, Culture, Sports, Science and Technology of Japan).

 

Prof. Michael R. Berthold Department of Computer and Information Science, University of Konstanz, Germany Homepage: http://www.inf.uni-konstanz.de/bioml/staff/berthold/index.html Title: Supporting Creativity: Towards Associative Discovery of New Insights   Co-Authors: Fabian Dill (University of Konstanz, Germany)                        Tobias Kötter (University of Konstanz, Germany)                        Kilian Thiel (University of Konstanz, Germany) Abstract: In this paper we outline an approach for network-based information access and exploration. In contrast to existing methods, the presented framework allows for the integration of both semantically meaningful information as well as loosely coupled information fragments from heterogeneous information repositories. The resulting Bisociative Information Networks (BisoNets) together with explorative navigation methods facilitate the discovery of links across diverse domains. In addition to such ``chains of evidence'', they enable the user to go back to the original information repository and investigate the origin of each link, ultimately resulting in the discovery of previously unknown connections between information entities of different domains, subsequently triggering new insights and supporting creative discoveries.   Short Biography: After receiving his PhD from Karlsruhe University, Germany Michael Berthold spent over seven years in the US, among others at Carnegie Mellon University, Intel Corporation, the University of California at Berkeley and - most recently - as director of an industrial think tank in South San Francisco. Since August 2003 he holds the Nycomed-Chair for Bioinformatics and Information Mining at Konstanz University, Germany where his research focuses on using machine learning methods for the interactive analysis of large information repositories in the Life Sciences. M. Berthold is Past President of the North American Fuzzy Information Processing Society, Associate Editor of several journals and a Vice President of the IEEE System, Man, and Cybernetics Society. He has been involved in the organization of various conferences, most notably the IDA-series of symposia on Intelligent Data Analysis and the conference series on Computational Life Science. Together with David Hand he co-edited the successful textbook "Intelligent Data Analysis: An Introduction" which has recently appeared in a completely revised, second edition.

 

Prof. Robert C. Holte Department of Computing Science, University of Alberta, Canada Homepage: http://www.cs.ualberta.ca/~holte/ Title: Cost-sensitive Classifier Evaluation using Cost Curves Co-Author: Chris Drummond (National Research Council, Ottawa) Abstract: The evaluation of classifier performance in a cost-sensitive setting is straightforward if the operating conditions (misclassification costs and class distributions) are fixed and known. When this is not the case, evaluation requires a method of visualizing classifier performance across the full range of possible operating conditions. This talk outlines the most important requirements for cost-sensitive classifier evaluation for machine learning and KDD researchers and practitioners, and introduces a recently developed technique for classifier performance visualization -- the cost curve -- that meets all these requirements. Short Biography: Dr. Robert Holte is a professor in the Computing Science Department of the University of Alberta. He is a well-known member of the international machine learning research community, former editor-in-chief of the "Machine Learning" journal, and past director of the Alberta Ingenuity Centre for Machine Learning (AICML). His current machine learning research focuses on learning in game-playing (for example: opponent modeling in poker, and the use of learning for gameplay analysis of commercial computer games). In addition to machine learning he undertakes research in single-agent search (pathfinding): in particular, the use of automatic abstraction techniques to speed up search. He has over 75 scientific papers to his credit, covering both pure and applied research, and has served on the organizing committees of numerous major international AI conferences, including being program co-chair for AAAI in 2007 (AAAI is the major international conference run by the Association for the Advancement of Artificial Intelligence), and co-founder of the SARA symposium series (SARA is the Symposium on Abstraction, Reformulation, and Approximation).

 

Prof. Genshiro Kitagawa Institute of Statistical Mathematics, Japan Homepage: http://tswww.ism.ac.jp/kitagawa/HTML-new0511/index.html Title: Prospective Scientific Methodology in Knowledge Society   Abstract: Due to the change of the society and the data environment of the scientific researches and the meaning of the knowledge, the role of scientific methodology is changing. We consider the possibility of establishing a new scientific base in coming knowledge society. In particular, we focus on the role of statistical science in knowledge society. Short Biography: Genshiro Kitagawa is Director General of the Institute of Statistical Mathematics, Executive Director of the Research Organization of Information and Systems and Professor of Statistical Science at the Graduate University for Advanced Study. His primary interests are in time series analysis, non-Gaussian nonlinear filtering, statistical modeling and discovery science. He is the executive editor of the Annals of the Institute of Statistical Mathematics, co-authors of Smoothness Priors Analysis of Time Series, Akaike Information Criterion Statistics, Information Criteria and Statistical Modeling, and several Japanese books. He was awarded the Japan Statistical Society Prize in 1997 and Ishikawa Prize in 1999, and is a Fellow of the American Statistical Association and an elected member of the International Statistical Institute. Currently he is the president of the Japan Statistical Society, chief director of the Japanese Federation of Statistical Science Associations, councilor of the International Statistical Institute and International Association for Statistical Computing.