DEFT2018 : recherche d'information et analyse de sentiments dans des tweets concernant les transports en Île de France

Abstract : Information Retrieval and Sentiment Analysis in Tweets about Public Transportation in Île de France Region This paper presents the 2018 DEFT text mining challenge. From a corpus of tweets, four tasks were proposed : first, to identify tweets about public transportation ; second, based on those tweets, to identify the global polarity (negative, neutral, positive, mixed), to identify clues of sentiment and target, and to annotate each tweet in terms of source and target concerning all expressed sentiments. Twelve teams participated, mainly on the two first tasks. On the identification of tweets about public transportation, micro F-measure values range from 0.827 to 0.908. On the identification of the global polarity, micro F-measure values range from 0.381 to 0.823.
Document type :
Conference papers
Complete list of metadatas

https://hal.archives-ouvertes.fr/hal-01839407
Contributor : Amel Fraisse <>
Submitted on : Friday, September 14, 2018 - 4:44:28 PM
Last modification on : Friday, May 3, 2019 - 4:56:47 PM
Long-term archiving on : Saturday, December 15, 2018 - 2:01:16 PM

File

deft2018-recherche-dinformatio...
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01839407, version 1

Citation

Patrick Paroubek, Cyril Grouin, Patrice Bellot, Vincent Claveau, Iris Eshkol-Taravella, et al.. DEFT2018 : recherche d'information et analyse de sentiments dans des tweets concernant les transports en Île de France. DEFT 2018 - 14ème atelier Défi Fouille de Texte, May 2018, Rennes, France. pp.1-11. ⟨hal-01839407⟩

Share

Metrics

Record views

634

Files downloads

182