Détection de communautés dans les grands graphes d'interactions (multiplexes) : état de l'art

Rushed Kanawati (1)

1 LIPN - Laboratoire d'Informatique de Paris-Nord

In this paper, we present a quick survey of current approaches for community detection in multiplex networks. Most of existing approaches are based on transforming, in one way or another, the problem into a problem of community detection in simple graphs. This can basically be made either by applying a slice aggregation strategy, either by applying community detection on each slice aside then merging obtained partitions using ensemble clustering like approaches. We argue that more interesting approaches would be based on exploring all slices at once. Recently proposed multiplex metrics allow to extend the application of promising community detection approaches, mainly grain-based approaches.

Mots clés: Détection de communautés; réseaux multiplexes; graphes d'interaction.; graphes d'interaction

Domaines: Intelligence artificielle [cs.AI]

Rushed Kanawati. Détection de communautés dans les grands graphes d'interactions (multiplexes) : état de l'art. 2013. ⟨hal-00881668v2⟩