The work presented in this thesis deals with data mining approaches for the analysis of attributed graphs. An attributed graph is a graph where properties, encoded by means of attributes, are associated to each vertex. In such data, our objective is the discovery of subgraphs formed by several dense groups of vertices that are homogeneous with respect to the attributes. More precisely, we define the constraint-based extraction of collections of subgraphs densely connected and such that the vertices share enough attributes. To this aim, we propose two new classes of patterns along with sound and complete algorithms to compute them efficiently using constraint-based approaches. The first family of patterns, named Maximal Homogeneous Clique Set (MHCS), contains patterns satisfying constraints on the number of dense subgraphs, on the size of these subgraphs, and on the number of shared attributes. The second class of patterns, named Collection of Homogeneous k-clique Percolated components (CoHoP), is based on a relaxed notion of density in order to handle missing values. Both approaches are used for the analysis of scientific collaboration networks and protein-protein interaction networks. The extracted patterns exhibit structures useful in a decision support process. Indeed, in a scientific collaboration network, the analysis of such structures might give hints to propose new collaborations between researchers working on the same subjects. In a protein-protein interaction network, the analysis of the extracted patterns can be used to study the relationships between modules of proteins involved in similar biological situations. The analysis of the performances, on real and synthetic data, with respect to different attributed graph characteristics, shows that the proposed approaches scale well for large datasets.

https://tel.archives-ouvertes.fr/tel-00858751/document

Finding homogeneous collections of dense subgraphs using constraint-based data mining approaches

Nous decrivons dans cet article notre participation a l'edition 2014 de DEFT. Nous nous interessons a la tâche consistant a associer des noms de session aux articles d'une conference. Pour ce faire, nous proposons une approche originale, symbolique et non supervisee, de decouverte de connaissances. L'approche combine des methodes de fouille de donnees sequentielles et de fouille de graphes. La fouille de sequences permet d'extraire des motifs frequents dans le but de construire des descriptions des articles et des sessions. Ces descriptions sont ensuite representees par un graphe. Une technique de fouille de graphes appliquee sur ce graphe permet d'obtenir des collections de sous-graphes homogenes, correspondant a des collections d'articles et de noms de sessions.

https://hal.archives-ouvertes.fr/hal-01113464/document

Fouille de données pour associer des noms de sessions aux articles scientifiques

Dans cet article, nous proposons une approche pour explorer des textes de taille importante en mettant en evidence des sous-parties coherentes. Cette methode d'exploration s'appuie sur une representation en graphe du texte, en utilisant le modele linguistique de Hoey pour selectionner et apparier les phrases dans le graphe. Notre contribution porte sur l'utilisation de techniques de fouille de graphes sous contraintes pour extraire des sous-parties pertinentes du texte (c'est-a-dire des collections de sous-reseaux phrastiques homogenes). Nous avons realise des experimentations sur deux textes anglais de taille consequente pour montrer l'interet de l'approche que nous proposons.

Fouille de graphes sous contraintes linguistiques pour l'exploration de grands textes (Graph Mining Under Linguistic Constraints to Explore Large Texts) [in French]

In this paper, we propose an approach to explore large texts by highlighting coherent sub-parts. The exploration method relies on a graph representation of the text according to Hoey's linguistic model which allows the selection and the binding of adjacent and non-adjacent sentences. The main contribution of our work consists in proposing a method based on both Hoey's linguistic model and a special graph mining technique, called CoHoP mining, to extract coherent sub-parts of the graph representation of the text. We have conducted some experiments on several English texts showing the interest of the proposed approach.

/pdf/graph-mining-under-linguistic-constraints-for-exploring-1ctqla7nki.pdf

Solen Quiniou

Papers

Fouille de graphes sous contraintes linguistiques pour l'exploration de grands textes (Graph Mining Under Linguistic Constraints to Explore Large Texts) [in French]

Graph Mining under Linguistic Constraints for Exploring Large Texts