Fabrice Guillet, Gilbert Ritschard, and Djamel Abdelkader Zighed (Eds.)

Advances in Knowledge Discovery and Management

Studies in Computational Intelligence,Volume 398

Editor-in-Chief

Prof. Janusz Kacprzyk

Systems Research Institute

Polish Academy of Sciences

ul. Newelska 6

01-447Warsaw

Poland

E-mail: kacprzyk@ibspan.waw.pl

Further volumes of this series can be found on our

homepage: springer.com

Vol. 378. János Fodor, Ryszard Klempous, and

Carmen Paz Suárez Araujo (Eds.)

Recent Advances in Intelligent Engineering Systems, 2011ISBN 978-3-642-23228-2

Vol. 379. Ferrante Neri, Carlos Cotta, and

Pablo Moscato (Eds.)

Handbook of Memetic Algorithms, 2011ISBN 978-3-642-23246-6

Vol. 380.Anthony Brabazon,Michael O’Neill, and

Dietmar Maringer (Eds.)

Natural Computing in Computational Finance, 2011ISBN 978-3-642-23335-7

Vol. 381. Rados law Katarzyniak, Tzu-Fu Chiu,

Chao-Fu Hong, and Ngoc Thanh Nguyen (Eds.)

Semantic Methods for Knowledge Management and

Communication, 2011ISBN 978-3-642-23417-0

Vol. 382. F.M.T. Brazier, Kees Nieuwenhuis, Gregor Pavlin,

Martijn Warnier, and Costin Badica (Eds.)

Intelligent Distributed Computing V, 2011

ISBN 978-3-642-24012-6

Vol. 383. Takayuki Ito,Minjie Zhang,Valentin Robu,

Shaheen Fatima, and Tokuro Matsuo (Eds.)

New Trends in Agent-Based Complex AutomatedNegotiations, 2012ISBN 978-3-642-24695-1

Vol. 384. DaphnaWeinshall, Jorn Anemuller,

and Luc van Gool (Eds.)

Detection and Identification of Rare Audiovisual Cues, 2012ISBN 978-3-642-24033-1

Vol. 385.Alex Graves

Supervised Sequence Labelling with Recurrent NeuralNetworks, 2012

ISBN 978-3-642-24796-5

Vol. 386.Marek R. Ogiela and Lakhmi C. Jain (Eds.)

Computational Intelligence Paradigms in Advanced Pattern

Classification, 2012ISBN 978-3-642-24048-5

Vol. 387. David Alejandro Pelta, Natalio Krasnogor,

Dan Dumitrescu, Camelia Chira, and Rodica Lung (Eds.)

Nature Inspired Cooperative Strategies for Optimization

(NICSO 2011), 2011ISBN 978-3-642-24093-5

Vol. 388. Tiansi Dong

Recognizing Variable Environments, 2012

ISBN 978-3-642-24057-7

Vol. 389. Patricia Melin

Modular Neural Networks and Type-2 Fuzzy Systems forPattern Recognition, 2012ISBN 978-3-642-24138-3

Vol. 390. Robert Bembenik, Lukasz Skonieczny,

Henryk Rybinski, and Marek Niezgodka (Eds.)

Intelligent Tools for Building a Scientific InformationPlatform, 2012

ISBN 978-3-642-24808-5

Vol. 391. Herwig Unger, Kyandoghere Kyamaky,

and Janusz Kacprzyk (Eds.)

Autonomous Systems: Developments and Trends, 2012ISBN 978-3-642-24805-4

Vol. 392. Narendra Chauhan,Machavaram Kartikeyan,

and Ankush Mittal

Soft Computing Methods for Microwave and Millimeter-Wave

Design Problems, 2012ISBN 978-3-642-25562-5

Vol. 393. Hung T. Nguyen,Vladik Kreinovich, BerlinWu,

and Gang Xiang

Computing Statistics under Interval and FuzzyUncertainty, 2012

ISBN 978-3-642-24904-4

Vol. 394. David A. Elizondo,Agusti Solanas,

and Antoni Martınez-Balleste (Eds.)

Computational Intelligence for Privacy

and Security, 2012ISBN 978-3-642-25236-5

Vol. 395. Srikanta Patnaik and Yeon-MoYang (Eds.)

Soft Computing Techniques in Vision Science, 2012ISBN 978-3-642-25506-9

Vol. 396.Marielba Zacarias and

Jose Valente de Oliveira (Eds.)

Human-Computer Interaction: The Agency Perspective, 2012ISBN 978-3-642-25690-5

Vol. 397. Elena Nikolaevskaya,Alexandr Khimich, and

Tamara Chistyakova

Programming with Multiple Precision, 2012ISBN 978-3-642-25672-1

Vol. 398. Fabrice Guillet, Gilbert Ritschard,

and Djamel Abdelkader Zighed (Eds.)

Advances in Knowledge Discovery and Management, 2012ISBN 978-3-642-25837-4

Fabrice Guillet, Gilbert Ritschard,and Djamel Abdelkader Zighed (Eds.)

Advances in KnowledgeDiscovery and Management

Volume 2

123

EditorsProf. Fabrice GuilletNantes UniversityFrance

Prof. Gilbert RitschardUniversite de GeneveSwitzerland

Prof. Djamel Abdelkader ZighedUniversite Lumiere Lyon 2

Bron

France

ISSN 1860-949X e-ISSN 1860-9503ISBN 978-3-642-25837-4 e-ISBN 978-3-642-25838-1DOI 10.1007/978-3-642-25838-1Springer Heidelberg New York Dordrecht London

Library of Congress Control Number: 2010928588

c© Springer-Verlag Berlin Heidelberg 2012

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or partof the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,recitation, broadcasting, reproduction onmicrofilms or in any other physical way, and transmission orinformation storage and retrieval, electronic adaptation, computer software, or by similar or dissimilarmethodology now known or hereafter developed. Exempted from this legal reservation are brief ex-cerpts in connection with reviews or scholarly analysis or material supplied specifically for the purposeof being entered and executed on a computer system, for exclusive use by the purchaser of the work.Duplication of this publication or parts thereof is permitted only under the provisions of the CopyrightLaw of the Publisher’s location, in its current version, and permission for use must always be obtainedfrom Springer. Permissions for use may be obtained through RightsLink at the Copyright ClearanceCenter.Violations are liable to prosecution under the respective Copyright Law.The use of general descriptive names, registered names, trademarks, service marks, etc. in this publi-cation does not imply, even in the absence of a specific statement, that such names are exempt from therelevant protective laws and regulations and therefore free for general use.While the advice and information in this book are believed to be true and accurate at the date ofpublication, neither the authors nor the editors nor the publisher can accept any legal responsibility forany errors or omissions that may be made. The publisher makes no warranty, express or implied, withrespect to the material contained herein.

Printed on acid-free paper

Springer is part of Springer Science+Business Media (www.springer.com)

Preface

During the last decade, the French-speaking scientific community developed a very

strong research activity in the field of Knowledge Discovery and Management

(KDM or EGC for “Extraction et Gestion des Connaissances” in French), which

is concerned with, among others, Data Mining, Knowledge Discovery, Business In-

telligence, Knowledge Engineering and Semantic Web.

This emerging research area has also been strongly stimulated by the rapid

growth of information systems and the web semantic issues. The success of the

first two French-speaking EGC Conferences, held in Nantes in 2001 and in Mont-

pellier in 2002, resulted naturally in 2003 in the foundation of the International

French speaking EGC Association1. Since, the Association yearly organizes con-

ferences and workshops with the aim of promoting exchanges between researchers

and companies concernedwith KDM and its application in business, administration,

industry or public organizations.

The recent and novel research contributions collected in this book are extended

and reworked versions of a selection of the best papers that were originally presented

in French at the EGC 2010 Conference held in Hammamet Tunisia, on January

2010. The 12 best papers that have been selected for this book are issued from

the 32 papers accepted in long format with a 23% acceptance ratio among the 139

papers submitted to the EGC2010 conference2.

Structure of the Book

The volume is organized in three parts.

1 Association “Extraction et Gestion des Connaissances” (EGC), see

http://www.egc.asso.fr2 For further details about EGC2010 conference see “10th International French-Speaking

Conference on Knowledge Discovery and Management (EGC2010): conference report”

by Sadok Ben Yahia and Jean-Marc Petit, in ACM SIGKDD Explorations, volume 12,

issue 1.

VI Preface

Part I – Various Aspects of Data Cube and Ontology-Based

Representations

In the first chapter, entitled Constrained Closed and Quotient Cubes, Rosine Cic-

chetti, Lotfi Lakhal and Sébastien Nedjar investigate reduced representations for the

Constrained Cube. They propose two representations which discard redundancies,

are information lossless and avoid to compute the whole data cubes: the Constrained

Closed and Quotient Cube. Provided with the former, the decision maker can per-

form OLAP classification and querying. The latter adds navigation within the cube

to the previous capabilities. Hence according to her/his needs, the user can choose

the more suitable representation.

In the second chapter, A New Concise and Exact Representation of Data Cubes,

Hanen Brahmi, Tarek Hamrouni, Riadh Ben Messaoud and Sadok Ben Yahia intro-

duce a new concise and exact representation called closed non derivable data cubes

(CND-Cube). It is based on the concept of non derivable minimal generators. They

also propose a novel algorithm dedicated to the mining of CND-Cube from mul-

tidimensional databases and investigate in detail the theoretical foundations of this

concise representation.Moreover, they discuss Rollup/Drilldown semantics over the

CND-Cube and validate their approach on a set of real and benchmark datasets.

Their experiment results show the effectiveness of the approach in comparison with

those fitting in the same trend.

Michel Buffa and Catherine Faron-Zucker are interested in Ontology-Based Ac-

cess Rights Management. They propose an approach to manage access rights in a

content management systems which relies on semantic web models and technolo-

gies. They developed the AMO ontology which consists in 1) a set of classes and

properties dedicated to the annotation of resources whose access should be con-

trolled and in 2) a base of inference rules modelling the access management strategy

to carry out. When applied to the annotations of the resources whose access should

be controlled, these rules enable to manage access according to a given strategy.

This modelisation is flexible, extendable and ensures the adaptability of the AMO

ontology to any access management strategy.

In the last chapter of this first part, Souhir Gahbiche, Nathalie Pernelle and Fatiha

Saïs addresse the problem of explanation of data reconciliation decisions that are ob-

tained by automatic numerical informed by ontology knowledge data reconciliation

methods. These methods may give rise to decision errors and to approximated re-

sults. They have developed an explanation model based on Coloured Petri Nets for-

malism. This model allows to generate a graphical and readable explanation, which

takes into account all the semantic knowledge that are involved in the similarity

computation of a reference pair.

Part II – Efficient Pattern Mining Issues

The first of chapter of this second part by Mehdi Khiari and Patrice Boizumault and

Bruno Crémilleux proposes to combine Constraint Programming and Constraint-

based Mining. By investigating the relationship between constraint-based mining

Preface VII

and constraint satisfaction problems, they propose a generic approach to model and

mine queries involving several local patterns (n-ary patterns). The resulting frame-

work for pattern set mining is very flexible and allows the user to easily express in

a declarative way a wide range of problems for a wide range of data mining tasks.

The next chapter written by Marc Boullé is concerned with Simultaneous Par-

titioning of Input and Class Variables for Supervised Classification Problems with

Many Classes. It extends discretization and value grouping methods, based on the

partitioning of both the input and class variables. The best joint partitioning is

searched by maximizing a Bayesian model selection criterion. This preprocessing

is exploited as a preparation for the naive Bayes classifier, and demonstrates high

performance in problems with hundreds of classes.

Lionel Martin and his co-authors describe an Interactive and progressive con-

straint definition for dimensionality reduction and visualization. They propose a

tool for semi-supervised, constraint-based projection of data. Starting from an ini-

tial linear projection, a user can specify constraints, in order to move away or closer

of some pairs of objects. Based on the Uzawa algorithm, a new projection is com-

puted that takes these constraints into account. The user can iteratively add new

constraints. The implementation is based on Explorer3D, the 3D projection tool de-

veloped at the LIFO (Computer Science Laboratory of Orleans, France).

The last chapter of this part by Anne Laurent and his co-authors deals with Ef-

ficient parallel mining of gradual patterns on multicore processors. Mining grad-

ual patterns plays a crucial role in many real world applications where huge vol-

umes of complex numerical data must be handled, e.g., biological databases, survey

databases, data streams or sensor readings. Gradual patterns highlight complex or-

der correlations of the form “The more/less X, the more/less Y”. They present an

efficient parallel algorithm to mine gradual patterns on multicore processors. They

experimentally show that this algorithm significantly improves the state of the art

and scales very well with the number of cores available.

Part III – Data Preprocessing and Information Retrieval

In the first one, entitled Analyzing Old Documents Using a Complex Approach:

application to lettrines indexing, byMickael Coustaty and his co-authors proposes a

methodology based on a complex approach to analyze and to characterize graphical

images extracted from old documents. Based on a comparison between historians

and computer science approaches, the authors propose a novel method to describe

images using a complex approach to globally describing an image with respect to

its structure, sense, and elements.

In Identifying relevant features of images from their 2-D topology, Marc Joliveau

introduces a new method for the visual perception problem that focuses on the in-

trinsic two dimensional topology of images to extract their principal features. Vali-

dations on three different datasets show that the few features extracted by the method

provide enough intelligible information to efficiently and fastly identify similarities

between the images of a database.

VIII Preface

The next chapter by Radja Messai et al. entitled Analyzing Health Consumer Ter-

minology for Query Reformulation Tasks describes the analysis and characterisation

of the health consumer terminology in the breast cancer field. The results have been

used in a pilot study to enhance the reformulation of health consumers’ queries. The

results have showed significant differences between the health consumer terminol-

ogy and the professional one. Such studies are important to provide health services

more adapted to the language and the level of knowledge of health consumers.

The last chapter, proposed by Patrick Bosc et al., addresses the plethoric answer

problem that often arises when end-users have an approximate idea of how to for-

mulate a query to retrieve what they are looking for from large-scale databases. A

possible approach to reduce the set of retrieved items and to make it more man-

ageable is to constrain the initial query with additional predicates. The approach

presented in this paper relies on the identication of correlation links between pre-

defined predicates related to attributes of the relation of interest. Thus, the initial

query is strengthened by additional predicates that are semantically close to the

user-specified ones.

Acknowledgments

The editors would like to thank the chapter authors for their insights and contribu-

tions to this book.

The editors would also like to acknowledge the members of the review commit-

tee and the associated referees for their involvement in the review process of the

book. Their in depth reviewing, criticisms and constructive remarks significantly

contributed to the high quality of the retained papers.

A special thank goes to Bruno Pinaud who has efficiently composed and laid out

the manuscript.

Finally, we thank Springer and the publishing team, and especially T. Ditzinger

and J. Kacprzyk, for their confidence in our project.

Nantes, Geneva, Lyon Fabrice Guillet

September 2011 Gilbert Ritschard

Djamel A. Zighed

Preface IX

Review Committee

All published chapters have been reviewed by at least 2 referees.

• Tomas Aluja (UPC, Barcelona, Spain)

• Nadir Belkhiter (Univ. Laval, Québec,

Canada)

• Sadok Ben Yahia (Univ. Tunis, Tunisia)

• Younès Bennani (Univ. Paris 13, France)

• Omar Boussaid (Univ. Lyon 2, France)

• Paula Brito (Univ. of Porto, Portugal)

• Francisco de A. T. De Carvalho (Univ.

Federal de Pernambuco, Brazil)

• Gilles Falquet (Univ. of Geneva, Switzer-

land)

• Jean-Gabriel Ganascia (Univ. Paris 6,

France)

• Pierre Gancarski (Univ. of Strasbourg,

France)

• Howard Hamilton (Univ. of Regina,

Canada)

• Robert Hilderman (Univ. of Regina,

Canada)

• Petra Kraemer (Univ. La Rochelle,

France)

• Philippe Lenca (Telecom Bretagne, Brest,

France)

• Philippe Leray (Univ. of Nantes, France)

• Stan Matwin (Univ. of Ottawa, Canada)

• Monique Noirhomme (FUNDP, Namur,

Belgium)

• Matthieu Perreira Da Silva (Univ. La

Rochelle, France)

• Vincent Pisetta (Univ. Lyon 2, France)

• Pascal Poncelet (LIRMM, Univ. of Mont-

pellier, France)

• Zbigniew Ras (Univ. of North Carolina,

USA)

• Jan Rauch (University of Prague, Czech

Republic)

• Chiara Renso (KDDLAB - ISTI CNR,

Italy)

• Lorenza Saitta (Univ. di Torino, Italy)

• Ansaf Salleb-Aouissi (Columbia Univ.,

New York, USA)

• Florence Sédes (Univ. of Toulouse 3,

France)

• Arno Siebes (Univ. Utrecht, The Nether-

lands)

• Dan Simovici (Univ. of Massachusetts

Boston, USA)

• Stefan Trausan-Matu (Univ. of Bucharest,

Romania)

• Rosanna Verde (Second Univ. of Naples,

Italy)

• Christel Vrain (Univ. of Orléans, France)

• Jef Wijsen (Univ. of Mons-Hainaut, Bel-

gium)

• Chongsheng Zhang (Univ. of Nice,

France)

Associated Reviewers

Yassine Benabbas,

Patrick Bosc,

Marc Boullé,

Hanen Brahmi,

Guillaume Cleuziou,

Mickaël Coustaty,

Bruno Cremilleux,

Catherine Faron Zucker,

Manfredotti,

Anne Laurent,

Lionel Martin,

Radja Messai,

Jonathan Weber

Manuscript Coordinator

Bruno Pinaud (Univ. of Bordeaux I, France)

Contents

Part I Data Cube and Ontology-Based Representations

Constrained Closed and Quotient Cubes . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

Rosine Cicchetti, Lotfi Lakhal, and Sébastien Nedjar

A New Concise and Exact Representation of Data Cubes . . . . . . . . . . . . . 27

Hanen Brahmi, Tarek Hamrouni, Riadh Ben Messaoud,

and Sadok Ben Yahia

Ontology-Based Access Rights Management . . . . . . . . . . . . . . . . . . . . . . . . 49

Michel Buffa and Catherine Faron-Zucker

Explaining Reference Reconciliation Decisions: A Coloured Petri Nets

Based Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

Souhir Gahbiche, Nathalie Pernelle, and Fatiha Saïs

Part II Efficient Pattern Mining

Combining Constraint Programming and Constraint-Based Mining

for Pattern Discovery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85

Mehdi Khiari, Patrice Boizumault, and Bruno Crémilleux

Simultaneous Partitioning of Input and Class Variables for Supervised

Classification Problems with Many Classes . . . . . . . . . . . . . . . . . . . . . . . . . 105

Marc Boullé

Interactive and Progressive Constraint Definition for Dimensionality

Reduction and Visualization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121

Lionel Martin, Matthieu Exbrayat, Guillaume Cleuziou,

and Frédéric Moal

Efficient Parallel Mining of Gradual Patterns on Multicore Processors . . 137

Anne Laurent, Benjamin Négrevergne, Nicolas Sicard,

and Alexandre Termier

XII Contents

Part III Data Preprocessing and Information Retrieval

Analyzing Old Documents Using a Complex Approach: Application

to Lettrines Indexing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155

Mickael Coustaty, Vincent Courboulay, and Jean-Marc Ogier

Identifying Relevant Features of Images from Their 2-D Topology . . . . . 173

Marc Joliveau

Analyzing Health Consumer Terminology for Query Reformulation

Tasks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191

Radja Messai, Michel Simonet, Nathalie Bricon-Souf,

and Mireille Mousseau

An Approach Based on Predicate Correlation to the Reduction of

Plethoric Answer Sets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213

Patrick Bosc, Allel Hadjali, Olivier Pivert, and Grégory Smits

List of Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235

Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243