Soft Constraints for Pattern Mining - Archive ouverte HAL Accéder directement au contenu
Article Dans Une Revue Journal of Intelligent Information Systems Année : 2015

Soft Constraints for Pattern Mining

Résumé

Constraint-based pattern discovery is at the core of numerous data mining tasks. Patterns are extracted with respect to a given set of constraints (frequency, closedness, size, etc). In practice, many constraints require threshold values whose choice is often arbitrary. This difficulty is even harder when several thresholds are required and have to be combined.Moreover, patterns barelymissing a threshold will not be extracted even if they may be relevant. The paper advocates the introduction of softness into the pattern discovery process. By using Constraint Programming, we propose efficient methods to relax threshold constraints as well as constraints involved in patterns such as the top-k patterns and the skypatterns. We show the relevance and the efficiency of our approach through a case study in chemoinformatics for discovering toxicophores.
Fichier principal
Vignette du fichier
RIACL-UGARTE-2013-1.pdf (12.65 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-01024695 , version 1 (16-07-2014)

Identifiants

Citer

Willy Ugarte Rojas, Patrice Boizumault, Samir Loudni, Bruno Crémilleux, Alban Lepailleur. Soft Constraints for Pattern Mining. Journal of Intelligent Information Systems, 2015, 44 (2), pp.193-221. ⟨10.1007/s10844-013-0281-4⟩. ⟨hal-01024695⟩
219 Consultations
93 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More