Tutorials

PAKDD-05 has three tutorials that are held on May 18. Please refer to the conference program for information in detail. TUTORIAL A - GRAPH MINING TECHNIQUES AND THEIR APPLICATIONS INSTRUCTORS Sharma Chakravarthy, University of Texas at Arlington, U.S.A TIME 9:00-12:00, May 18th 2005 ABSTRACT In this tutorial, we argue that graph mining techniques are extremely important and very little attention has been paid to this research and technology so far. Most of the currently used mining approaches assume transactional and other forms of data. However, there are a large number of applications where the relationships among data objects are extremely important. For those applications, use of conventional approaches results in loss of information that will critically affect the knowledge that is discovered using earlier approaches. Mining techniques that preserve and exploit the domain characteristics is extremely important and graph mining is one such general purpose technique as it uses arbitrary graph representation using which complex relationships can be represented. In this tutorial, we overview conventional mining techniques, contrast them with the requirements of applications and introduce graph mining as an alternative approach for a large class of applications. We present details of several graph mining approaches along with their advantages and disadvantages. We then present a few applications that have used graph mining techniques beneficially. BIOGRAPHY Sharma Chakravarthy is Professor of Computer and Engineering Department at The University of Texas at Arlington, Texas. He is well known for his work on semantic query optimization, multiple query optimization, active databases (HiPAC project at CCA and Sentinel project at the University of Florida, Gainesville), and more recently scalability issues in graph mining and its applications. His group at UTA is currently developing DB-Subdue - a scalable system for graph mining, and InfoSift - a classification system for text, email, and web that uses graph mining techniques. His current research includes web technologies, stream data processing, mining and knowledge discovery - association, graph and text, active and real-time databases, distributed and heterogeneous databases, query optimization, and multi-media databases. He has published over 110 papers in refereed international journals and conference proceedings. TUTORIAL B - ROUGH SET APPROACH TO KDD In conjunction with Workshop B "Rough Set Techniques in Knowledge Discovery". INSTRUCTORS Hung Son Nguyen and Marcin Szczuka, Warsaw University, POLAND TIME 9:00-12:00, May 18th 2005 ABSTRACT In recent years, a growth of interest in rough set theory and its applications can be seen in the number of research papers submitted to international workshops, conferences, journals and edited books, including two main biannual conferences on rough sets and the special sub-line of LNCS series. A large number of efficient applications of rough sets in Knowledge Discovery from various types of databases have been developed. Rough sets are applied in many domains, such as, for instance, medicine, finance, marketing, telecommunication, conflict resolution, text mining, intelligent agents, image analysis, pattern recognition, bioinformatics, etc. This tutorial is intended to fulfill the needs of many researchers to understand the rough set methodologies to mining of standard and nonstandard data. The methodology based on rough sets can serve as a useful tool to complement capabilities of other data mining methods. The tutorial should help the audience to find out if some of the presented methods may support their own KDD&DM research. BIOGRAPHY Dr. Hung Son Nguyen is an assistant professor at the Faculty of Mathematics, Informatics and Mechanics, Warsaw University, Poland. He worked as a visiting scientist in Sweden and Northern Ireland. Dr. Hung Son Nguyen is an author and coauthor of more than 60 articles published in journals edited books and conference proceedings. His main research interest include: rough set theory and applications in data mining, hybrid and multi-layered learning methods, KDD with applications in text mining, image processing and bioinformatics. Dr. Marcin Szczuka is an assistant professor at the Faculty of Mathematics, Informatics and Mechanics, Warsaw University, Poland. He worked as a researcher and lecturer in Poland, Sweden and Japan. In last few years he published several research articles, edited some books and proceedings as well as prepared several international workshops and conferences. His main include: hybrid classification methods, knowledge discovery, compound and multi-level classification schemes, data mining, rough sets. TUTORIAL C - ADVANCED TECHNIQUES FOR INFORMATION AND IMAGE CLASSIFICATION FOR KNOWLEDGE MANAGEMENT AND DECISION MAKING INSTRUCTORS Parag Kulkarni, Capsilon India, Pune, INDIA TIME 13:30-16:30 May 18th 2005 ABSTRACT Many great companies failed because they could not classify their problems. Some companies succeeded with proper classification of problems but decisions taken were not correct. Basic step in problem solving is classifying the problem. With better classification, we can handle problems in focused and well-directed manner. But classifying problems, texts, images, people, etc are very complex tasks. Complexity is based on problems. In most of the cases non-linearity and complexity of these problems makes it very difficult to solve them. All the knowledge and information in the company or any institution needs to be classified to make it available whenever it is necessary. Here in this tutorial we will discuss about classification issues and various ways to address these problems. Many statistical techniques (like Bayesian, Markov Decision Process, SMDP, time-series method, probabilistic theories), Neural networks, etc are used for classification with very good success. Still for certain domain of problems where non-linearity is very high, this problem solving remains a challenging affair. In the process we will focus on various advanced classification techniques, which has potential to classify some of the most complex classification issues. BIOGRAPHY Dr. Parag is Ph.D. from computer department of IIT, Kharagpur. He is working in IT industry for more than 14 years. He is on research panel and Ph.D. guide for University of Pune, BITS and Symbiosis deemed University. He is member of IASTED technical committee of Parallel and Distributed Computing. Presently he is Research Head at Capsilon India, Pune. He is also Honorary Professor at AISSM Engineering College, Pune. He has worked as Senior Manager (R&D) at Scientific Applications Center, Siemens information systems Ltd., Pune. His areas of interest include image processing, AI, classification techniques, e-governance, security systems, decision systems, expert systems, load balancing and distributed computing. Click here to see the previous details about the call for tutorials. [ HOME ]