ICFCA 2007

Clermont-Ferrand, France, February 12-16, 2007

Invited Speakers

Dr. Nathalie Caspard - Pr. Michel Habib - Pr. Bernhard Ganter
Pr. Ali Jaoua - Dr. Engelbert Mephu Nguifo - Pr. Rokia Missaoui
Dr. Julio Valdes

[Dr. Nathalie Caspard Photo] Dr. Nathalie Caspard

Dr. Nathalie Caspard received her PhD in Computer Science from University Paris 1 and EHESS (Ecole des Hautes Etudes en Sciences Sociales) in 1998. Then, since 2000, she has been Associate Professor and a member of the LACL laboratory of the Faculty of Sciences and Technology of University Paris 12 Val-de-Marne. Her early research areas particularly dealt with the class of lattices obtained by "convex sets doublings", a generalisation of the interval doubling defined by Day in 1970 and later developed by himself and Geyer. Since then, she has been working especially on two classes of lattices : Coxeter lattices, which generalize the Permutohedron and connect her work with the theory of Coxeter groups, and lattices associated with closure systems, which relate to the relational model of databases.

[Pr. Michel Habib Photo] Pr. Michel Habib

Talk: Which graph algorithms for searching the Web ?
In 1999, independantly J. Kleinberg with the algorithm Hits and L. Page with PageRank, proposed ranking algorithms uniquely based on the graph structure of the Web graph (i.e. the structure of the links between the html pages). These algorithms are used by the search engines to sort the answers of a given query. The sucess story of Google popularized this approach which can be called "syntagraphically" (no use of semantics of the contents of the documents). We present what is known about the Web graph, and how these algorithms have been implemented and then examine their limits. Then we discuss the two major drawbacks of these algorithms: bombing and spamming. Can we find good graph algorithms to solve these problems in a "syntagraphically" way ?

[Pr. Bernhard Ganter photo] Pr. Bernhard Ganter

Bernhard Ganter Bernhard Ganter received his PhD in 1974 from the University of Darmstadt, Germany. He became a Professor of mathematics in 1978, and a Full Professor at Dresden University in 1993. Presently he is the head of the Algebra Institute in the Math department. Originally, his research interests were in discrete mathematics and in universal algebra. Around 1980, he was a member of the research group around Rudolf Wille that invented Formal Concept Analysis. This field has since then turned out to be useful for the formal treatment of data and knowledge.

Personal home page of Pr. Bernhard Ganter

[Pr. Ali Jaoua Photo] Pr. Ali Jaoua

Professor Ali Jaoua is a faculty member of computer Science and engineering department, University of Qatar, since 2000, in leave from University of Manar (Tunisia), where he supervised the Research Group on Algorithmic and Heuristics (URPAH: 50 Members). He obtained the degree of "Docteur es-Science" (1987) from the University Paul Sabatier of Toulouse (France), "Doctor Engineer" (1979) from Institute Polytechnic of Toulouse, and "Engineer in Computer Science" from ENSEEIHT of Toulouse (1977). His research areas are Software and Information Engineering. He has been working on the Use of Relational Methods in Computer Science since 1984, and the application of formal concept analysis in Information Engineering since 1990. He has been Invited Prof./researcher in the following universities: University Joseph Fournier, Grenoble, France, July 2004, University of Sherbrooke (Canada), in 1989, Sophia-Antipolis (I3S) (France) during December 1991. He has been associate Professor in computer science in Laval University (Quebec, Canada), 1989-1992. He has been invited speaker for some international conferences and universities during the last ten years. He also has organized several conferences in computer science, and has been the seminar coordinator for about 12 years of his carrier. He supervised about 12 Phd-thesis defended mostly on conceptual information engineering, and several Masters. He published about 35 papers in International Journals, and many conferences, and contributes in several books. He is a Member of the Steering Committee of Relational Methods in Computer Science Group, RELMICS, and Editorial member of the associated electronic journal (JORMICS), Member of the Editorial Board of the Journal of Computing, since 2004. Member of The International Reviewers Committee of the Journal of Web Engineering, since 2005. He has also been member of many associations in computer science.

[Dr. Engelbert Mephu Nguifo Photo] Dr. Engelbert Mephu Nguifo

Engelbert Mephu Nguifo received his Ph.D. degree in computer science from the University of Montpellier in 1993. In 1994 he became an assistant professor of computer science at University Institute of Technology of Lens (IUT of Lens - University of Artois). Since 2002, he is an associate professor. As member of the computer science research center of Lens (CRIL-CNRS) since 1994, his research interests focus on data mining techniques for intelligence applications, bioinformatics, and so, and especially on developing concept lattice-based data mining techniques. He has served as guest-editor for several journals and has served on the program committees of several international conferences. He has co-chaired many workshops on Formal Concept Analysis, and as well as the 4th international conference on Concept Lattices and their Applications (CLA'2006). He has organized the first workshop on concept lattices and data mining (CLKDD'2001) at Stanford University, and has served as conference chair of the third ICFCA held in Lens in 2005. He was head of the computer science department at IUT of Lens from 1998 to 2004. He is member of the ACM society.

Talk: Closing the Gap : Data mining meets Formal Concept Analysis
Still in its early teens, data mining (DM) aims to deal with huge databases and therefore with high dimentionality. Recent developments in DM show an increasing number of research works based on closure operators, that efficiently reduce such high dimensionality without information loss. Meanwhile since two decades, formal concept analysis (FCA) which is now a mature field, has built solid mathematical theories, methods and tools to favor data understanding. Many works in one discipline are being developed without deep connexion to works in the other. During this talk, i will show that DM and FCA goals are overlaped. I will describe and discuss recent developments that put FCA at the heart of DM. Finally i will describe our recent contributions that connected both domains, in two directions : building kernel knowledge for DM and, visualizing huge set of association rules.

[Pr. Rokia Missaoui Photo] Pr. Rokia Missaoui

Rokia Missaoui received her Ph.D. degree in Computer Science from Université de Montréal in 1988. She has been a full Professor in the Department of Computer Science and Engineering at UQO (Université du Québec en Outaouais) since August 2002. Before joining UQO, she was a professor at UQAM (Université du Québec à Montréal) between 1987 and 2002.

While at the beginning of her career she was concerned with performance and design issues in databases, her current research interests include data mining, formal concept analysis (mainly concept lattice construction and manipulation, and association rule mining), data warehousing, as well as content-based image retrieval and mining.

Personal home page of Pr. Rokia Missaoui

Talk: Algebraic Operators for Querying Concept Lattices
In data mining applications, users tend to be overwhelmed by the generated knowledge, even when the input is relatively small. This is particularly the case in association rule mining and concept lattice construction and manipulation. One way to handle this problem consists to define and implement operators that help the user query either the input data or the data mining result for targeted knowledge. In order to manipulate lattices derived from formal concept analysis (FCA) in a similar way as in relational databases or in a meta-rule guided mining way, we take a joint database-FCA perspective by using operators similar in spirit to relational algebra operators (e.g., selection, projection and join) and exploiting existing work related to operations on contexts and concept lattices to formalize such operators.

[Dr. Julio Valdez] Dr. Julio Valdes

Dr. Julio J. Valdés has a PhD in mathematics (1987). His areas of interest are: artificial intelligence (mathematical foundations of uncertainty processing and inexact reasoning), computational intelligence (fuzzy logic, neural networks, evolutionary algorithms, probabilistic reasoning, rough sets), data mining, hybrid systems, virtual reality, pattern recognition and image and signal processing. He also graduated in geophysics (1977), oriented to geomathematics, mathematical modeling of natural processes, data analysis of earth and environmental data, remote sensing, physics and chemistry of external geodynamic processes and geophysical prospecting. He has more than 150 publications in journal and scientific conferences and is the Chair of the IEEE Computational Intelligence Society, Ottawa Chapter. Currently he is Senior Research Officer at the National Research Council Canada, Institute for Information Technology.

Talk: Visual Data Mining with Virtual Reality Spaces