Fast Mining of Iceberg Lattices: A Modular Approach Using Generators - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Conference Papers Year : 2011

Fast Mining of Iceberg Lattices: A Modular Approach Using Generators

Petko Valtchev
Amedeo Napoli
Robert Godin
  • Function : Author
  • PersonId : 854703
Alix Boc
  • Function : Author
  • PersonId : 913853
Vladimir Makarenkov
  • Function : Author
  • PersonId : 913854


Beside its central place in FCA, the task of constructing the concept lattice, i.e., concepts plus Hasse diagram, has attracted some interest within the data mining (DM) field, primarily to support the mining of association rule bases. Yet most FCA algorithms do not pass the scalability test fundamental in DM. We are interested in the iceberg part of the lattice, alias the frequent closed itemsets (FCIs) plus precedence, augmented with the respective generators (FGs) as these provide the starting point for nearly all known bases. Here, we investigate a modular approach that follows a workflow of individual tasks that diverges from what is currently practiced. A straightforward instantiation thereof, Snow-Touch, is presented that combines past contributions of ours, Touch for FCIs/FGs and Snow for precedence. A performance comparison of Snow-Touch to its closest competitor, Charm-L, indicates that in the specific case of dense data, the modularity overhead is offset by the speed gain of the new task order. To demonstrate our method's usefulness, we report first results of a genome data analysis application.
Fichier principal
Vignette du fichier
cla2011.pdf (234.54 Ko) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-00640898 , version 1 (14-11-2011)


  • HAL Id : hal-00640898 , version 1


Laszlo Szathmary, Petko Valtchev, Amedeo Napoli, Robert Godin, Alix Boc, et al.. Fast Mining of Iceberg Lattices: A Modular Approach Using Generators. The Eighth International Conference on Concept Lattices and their Applications - CLA 2011, INRIA Nancy Grand Est - LORIA, Oct 2011, Nancy, France. ⟨hal-00640898⟩
244 View
167 Download


Gmail Facebook X LinkedIn More