Building a treebank of noisy user-generated content: The French Social Media Bank

Djamé Seddah (1, 2) , Benoît Sagot (2) , Marie Candito (2) , Virginie Mouilleron (2) , Vanessa Combet (2)

1 UP4 - Université Paris-Sorbonne
2 ALPAGE - Analyse Linguistique Profonde à Grande Echelle ; Large-scale deep linguistic processing

We introduce the French Social Media Bank, the first user-generated content treebank for French. Its first release contains 1,700 sentences from various Web 2.0 and social media sources (FACEBOOK, TWITTER, web forums), including data specifically chosen for their high noisiness.

Mots clés: Social Media; Treebanking; User Generated Content; Parsing; Social Media.

Domaines: Traitement du texte et du document

Djamé Seddah, Benoît Sagot, Marie Candito, Virginie Mouilleron, Vanessa Combet. Building a treebank of noisy user-generated content: The French Social Media Bank. TLT 11 - The 11th International Workshop on Treebanks and Linguistic Theories, Nov 2012, Lisbonne, Portugal. ⟨hal-00780898⟩