Fusion de préférences pour la détection de communautés dans les réseaux sociaux

Abstract : This article addresses the issue of preference fusion in the context of social networks. The aim of this work is to compute the collective preferences of a user group belonging to a given community from the individual preferences of each of them. In particular, we are interested here in the representation of individual preferences taking into account the uncertainty phenomenon. The uncertainty representing a source of complexity in decision making, we propose an effective preference fusion method based on belief function theory. This method allows to qualify some user communities by analyzing the collective preferences obtained during the fusion process. Specifically, it allows the characterization of communities sharing the same collective preferences. Experiments on generated data highlight the interest of the proposed solution.
Complete list of metadatas

Cited literature [13 references]  Display  Hide  Download

https://hal.inria.fr/hal-01270169
Contributor : Tassadit Bouadi <>
Submitted on : Friday, February 5, 2016 - 5:56:01 PM
Last modification on : Thursday, March 7, 2019 - 11:34:08 AM
Long-term archiving on : Saturday, November 12, 2016 - 11:09:37 AM

File

F.ELARBI_LFA2015.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01270169, version 1

Citation

Faten Elarbi, Tassadit Bouadi, Arnaud Martin, Boutheina Ben Yaghlane. Fusion de préférences pour la détection de communautés dans les réseaux sociaux. 24ème Conférence sur la Logique Floue et ses Applications, Nov 2015, Poitiers, France. ⟨hal-01270169⟩

Share

Metrics

Record views

551

Files downloads

210