Learning from Social Media and Contextualisation

Josiane Mothe (1, 2, 3)

1 IRIT-SIG - Systèmes d’Informations Généralisées
2 ESPE Toulouse - École supérieure du professorat et de l'éducation - Toulouse Midi-Pyrénées
3 UT2J - Université Toulouse - Jean Jaurès

Social media can be a rich source of information either to extract some trends (models)or peculiarities (weak signals). We focused in this talk on early depression detection from social media posts using machine learning techniques and presented some results. We also proposed to use the same type of model to detect and extract locations from short posts when user localisation is not available. Finally, we mentioned our current work on tweet contextualisation that aims helping users to understand short texts.

Mots clés: Social media

Domaines: Recherche d'information [cs.IR]

Josiane Mothe. Learning from Social Media and Contextualisation. Dagstuhl Seminar 17301 (2017), Jul 2017, Dagstuhl, Germany. pp.121-121. ⟨hal-02348226⟩