A Time and Space Routing Game Model applied to Visibility Competition on Online Social Networks

Abstract : Companies or users that want to do advertising on Online Social Networks, need to know in which News Feed it would be more efficient to post, at what time of day the posts would have more visibility and which topic of message they should choose to reach popularity. We propose to answer these questions with the help of the Routing Game Theory, which considers a finite number of players and splittable demands. First, we propose a visibility measure on a News Feed. Next, we present our game based on visibility measures, which is similar to the Weighted Allocation Game. Following this consideration, we provide the uniqueness of the Nash Equilibrium, a characterization via concave programming, and its closed form. Then, we propose a two time scale decentralized algorithm where each user estimates the total flow of messages in each News Feed and uses stochastic gradient algorithm to update their own flows. Finally, we use a real data set in order to estimate the parameters of our model. The goal of this model is to provide to the Online Social Network Authority a better knowledge of posting behavior of competitive users.
Type de document :
Communication dans un congrès
7th International Conference on NETwork Games COntrol and OPtimization (NETGCOOP 2014), Oct 2014, Trento, Italy. 2014
Liste complète des métadonnées

Littérature citée [31 références]  Voir  Masquer  Télécharger

https://hal.inria.fr/hal-01066455
Contributeur : Eitan Altman <>
Soumis le : samedi 20 septembre 2014 - 14:30:07
Dernière modification le : jeudi 11 janvier 2018 - 16:57:57
Document(s) archivé(s) le : vendredi 14 avril 2017 - 13:09:46

Fichier

alex-netgcoop.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : hal-01066455, version 1

Collections

Citation

Alexandre Reiffers-Masson, Eitan Altman, Yezekael Hayel. A Time and Space Routing Game Model applied to Visibility Competition on Online Social Networks. 7th International Conference on NETwork Games COntrol and OPtimization (NETGCOOP 2014), Oct 2014, Trento, Italy. 2014. 〈hal-01066455〉

Partager

Métriques

Consultations de la notice

287

Téléchargements de fichiers

264