Synasc 2005 - Invited talks

Invited talks

Natural Computation for Business Intelligence from Web Usage Mining

Ajith Abraham
IITA Professorship Program
School of Computer Science and Engineering,
Chung-Ang University, S. Korea

Abstract: The rapid e-commerce growth has made both business community and customers face a new situation. Due to intense competition on the one hand and the customer's option to choose from several alternatives, the business community has realized the necessity of intelligent marketing strategies and relationship management. Web usage mining attempts to discover useful knowledge from the secondary data obtained from the interactions of the users with the Web. Web usage mining has become very critical for effective Web site management, creating adaptive Web sites, business and support services,personalization, network traffic flow analysis and so on. This talk presents the important concepts of Web usage mining and its various practical applications. Further a web usage mining framework based on novel natural computation techniques is presented. Proposed framework is compared with several clustering and function approximation techniques like neural networks, fuzzy clustering, genetic programming and Takagi - Sugeno fuzzy inference system etc. The results are graphically illustrated and the practical significance is discussed in detail. Empirical results clearly show that the proposed Web usage-mining framework is efficient.

[1] Abraham A., World Wide Web Usage Mining, Computationally Intelligent Hybrid Systems: The Fusion of Soft Computing and Hard Computing, Seppo Ovaska (Ed.), John Wiley & Sons Inc. and IEEE Press,
ISBN 0471476684, Chapter 11, pp. 363 -396, 2004.
[2] Abraham A., Business Intelligence from Web Usage Mining, Journal of Information & Knowledge Management (JIKM), World Scientific Publishing Co., Singapore, Vol. 2, No. 4, pp. 375-390, 2003.
[3] Abraham A., i-Miner: A Web Usage Mining Framework Using Hierarchical Intelligent Systems, The IEEE International Conference on Fuzzy Systems, FUZZ-IEEE'03, IEEE Press, pp. 1129-1134, 2003.
[4] Abraham A. and Ramos V., Web Usage Mining Using Artificial Ant Colony Clustering and Genetic Programming, 2003 IEEE Congress on Evolutionary Computation (CEC2003), Australia, IEEE Press, pp.
1384-1391, 2003.

Semantics of Membrane Computing

Gabriel Ciobanu
Institute of Theoretical Informatics, Iasi, Romania
Institute e-Austria, Timisoara, Romania

Symbolic Computing Grid for Computational Origami

Tetsuo Ida
Department of Computer Science
University of Tsukuba
Tsukuba 305-8573, Japan

Abstract: Although grid technologies are coming closer to scientists and engineers in many domains of sciences and technologies, they have not reached  the level of maturity that the Internet provides to us in our daily life.   In this talk I will introduce Computation Origami System (to be abbreviated to COS hereafter)  as a case  for the necessity of a grid for symbolic computation. COS, which is under development by us, is a 'paper' folding system, where origami papers and folding operations are simulated by a computer. Algorithmically, paper folds by COS are formulated as symbolic constraint solving. Namely, a paper fold problem is modeled symbolically as a set of polynomial equalities that represent the constraints.  Although for visualizing origami we need to resort to traditional numeric computing to find the solutions of the set of polynomial equalities, we have a new aspect of computing: the formulation as symbolic constraint solving leads to the possibility of directly proving the correctness of origami constructions. This is achieved by interaction of COS and theorem prover Theorema.  We will show that the interaction of COS with Theorema will indeed solve a very sophisticated problem of constructing and proving  Morley's triangle by origami.  We will further point out that the proof requires heavy use of computer resources,  which can best be provided by a symbolic computing grid. To obtain a correct proof, we need to have a connection with Theorema server for Gr"obner bases computation for quite a long time and many times for finding out the optimal specification of parameters to the proof computation by Theorema.  By illustration I will show that these requirements are  fulfilled by a service oriented symbolic computing grid.

Membrane computing. Basic ideas, results, applications

Gheorghe Paun
Romanian Academy of Sciences, Bucuresti, Romania
Sevillia University, Spain

Abstract: Membrane computing is a branch of natural computing whose goal is to abstract computing models from the cell structure and functioning. In the basic model, one processes multisets, mainly by rewriting-like rules, in a compartmental stucture defined by hierarchically arranged membranes. There also are tissue-like and neural-models. All these models (called P systems) are distributed parallel computing devices. There were introduced many types of P systems. Most of them were proven to be computationally complete, while several classes of P systems with additional possibilities of producing an exponential working space can solve computationally hard problems in a feasible time (by a time-space trade-off). In the last years, membrane computing was also proven to be a promising framework for applications in various areas, especially in biology (but also in computer science, linguistics, economics, etc). The talk will present the basic ideas of membrane computing, the types of results (universality and efficiency), some research directions, and will briefly discuss some recent applications.

 Role of Knowledge/Annotations in  Complex KDD Processes.
Illustration in Web Usage Mining

Brigitte Trousse

INRIA Sophia Antipolis,  France

Abstract: the aim of this talk will be to address issues related to the role of knowledge and/or annotations in complex KDD processes.  We first introduce the notion of complexity in Knowledge Discovery from Databases and in Data Mining.  Two domains will be considered  in this talk: the analyst domain and  the analysed domain of expertise. We will  present a synthesis of research works issued from  Semantic Web and Knowledge Engineering in order to improve KDD processes or to support  the reuse of such processes. Then we will focus mainly on the importance of  capturing some knowledge in KDD and annotating such  KDD processes in order to keep the sense of the main decision-makings and to support the reuse of such processes. We will illustrate it in the context of Web Usage Mining.