Skip to Main content Skip to Navigation
Journal articles

Retweeter ou ne pas retweeter : Le dilemme des portails de diffusion d’information temps-réel

Abstract : The study of contextual features has been widely discussed in Information Retrieval (IR), but concrete applications on real data streams are not common. In this paper, we aim at doing retweet recommandation. Considering a user interest, we introduce a model to perform real-time online filtering of the Twitter stream using several contextual features. The model separates content and contextual aspects, achieving a very high velocity. Experiments were performed on the TREC Microblog 2015 and TREC 2016 Real-Time Summarization frameworks. Results show that our model is very efficient as well as effective on the 2015 collection. However, the results regarding effectiveness have not been confirmed on the 2016 framework. This led us to conduct a detailed analysis of the results with regard to our approach and the evaluation framework. This analysis showed an evaluation bias discussed at the end of the article.
Complete list of metadata

Cited literature [25 references]  Display  Hide  Download
Contributor : Open Archive Toulouse Archive Ouverte (oatao) Connect in order to contact the contributor
Submitted on : Tuesday, March 12, 2019 - 9:39:13 AM
Last modification on : Tuesday, October 19, 2021 - 2:24:18 PM
Long-term archiving on: : Thursday, June 13, 2019 - 1:43:43 PM


Files produced by the author(s)



Thomas Palmer, Gilles Hubert, Karen Pinel-Sauvagnat. Retweeter ou ne pas retweeter : Le dilemme des portails de diffusion d’information temps-réel. Document Numérique, Lavoisier, 2018, 21 (3), pp.81-103. ⟨10.3166/DN.21.3.81-103⟩. ⟨hal-02064611⟩



Record views


Files downloads