Contact : Chloé Clavel
In the context of the Social Computing team, we aimed at interconnecting complex (e.g. psychological or sociolinguistic) theories of the social sciences with the formalism underlying machine or deep learning models for socio-emotional phenomena such as sentiments, emotions, stances, cohesion, trust or self-confidence, with a focus on language modality. The envisaged applications are conversational systems and social network analysis for different domains: Education, Health or Human Resources.
Examples of research carried out in Education:
- The ANIMATAS (Innovative Training Network) project: we work on the development of social robots in education, proposing neural models for the analysis of people’s self-confidence in an interaction.
- The ANR project REVITALISE: we work on interactive systems for training students in public speaking.
Keywords : sentiment analysis, conversational artificial intelligence, human-agent interaction, natural language generation, natural language understanding, emotion recognition in conversations, language models, multimodality, social robotics
The topic is transdisciplinary, and is linked to the activities of the Economic and Social Sciences department of Telecom-Paris, in particular: