Augmenter les retweets sur Twitter : comment tirer parti des mentions ?

Abstract : While Twitter has become one of the most influential micro-blogging systems, the propagation of tweets or hashtags is still widely misunderstood. Information propagation in Twitter is mainly due to 'retweets' and 'mentions' but, while retweets only reach the circle of a source, mentions allows to spread an information far beyond its neighborhood in just one step. Studies show that mentions are not only widely used by Twitter users, but are fundamental in the popularity of tweets or hashtags. Solutions to help in mentioning the proper users could therefore give maximal exposure to a given tweet. In this paper, we propose a real time mention recommendation system to enhance the popularity of tweets by a strategic use of the mention utility. This system is based on a model of tweet propagation in a multiplex network whose basic bricks are supported by the study of a real dataset and that allows to make a clear difference between retweets due to mentions and other factors. Simulations of the model show a nice agreement with the empirical tweet popularity observed in the dataset and are further supported by analytical results. Using all these results, we propose an effective mention recommendation strategy and implement it in a Twitter Application.
Complete list of metadatas

Cited literature [18 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-01345815
Contributor : Lionel Tabourier <>
Submitted on : Friday, July 15, 2016 - 8:19:17 PM
Last modification on : Thursday, March 21, 2019 - 1:09:23 PM

File

pramanik_marami_2015.pdf
Files produced by the author(s)

Identifiers

Citation

Soumajit Pramanik, Qinna Wang, Maximilien Danisch, Mohit Sharma, Sumanth Bandi, et al.. Augmenter les retweets sur Twitter : comment tirer parti des mentions ?. Sixième conférence Modèles et Analyses Réseau : Approches Mathématiques et Informatique (MARAMI 2015), Oct 2015, Nîmes, France. ⟨10.3166/RIA.na.1-13⟩. ⟨hal-01345815⟩

Share

Metrics

Record views

175

Files downloads

177