This PhD thesis work belongs to Artificial Intelligence and Human-Machine Communication. The goal is to define reasoning models that support the design and the implementation of intelligent conversational agents that are capable of reasoning about the mental state of their interlocutor. This will lead to a more natural and engaging interaction. The originality of this subject is to propose a model that reasons not only on the dialogical goals in relation to the user's task, but also about the user's mental state (and especially its emotions) and the representation that the user has of the personality of the virtual agent, based on his/her dialogue selection and the affective signal it expresses. The agent is then capable to use the affective modality, in complement to the dialogical one, to better satisfy the user in the context of a commercial application.
Context
This research work is done in the frame of a collaboration between LIMSI-CNRS (team CPU) and the company DAVI, specialist of virtual character's animation and conversational agents. The DAVI system recently integrated the dialogical dimension in its animation model and the company is now interested in taking into account the affective dimension. Meanwhile, the CPU team at LIMSI-CNRS has been leading research in affective computing for the last 10 years, especially in emotion expression for virtual agents and automated reasoning about affects in the context of social interactions.
Objectives
The goal of this PhD work is to combine three dimensions that have been considered separately until now:
- Reasoning (using logical models such as BDI) about the interlocutor's mental state and task-related goals;
- Modeling the dialogical goals, which are not limited to the task, but also include meta-goals such as expressing a personality or having a pleasant dialogue;
- Representing and reasoning about affects' expression in the context of communication (e.g. intentional smiling to express sympathy)
Work program
M0-M6 : bibliography on these 3 research questions
M6-M12 : first reasoning model
M12-M18 : implementation
M6-M18 : definition of the evaluation methodology
M18-M22 : first evaluation
M22-M30 : integration of a knowledge representation model
M30-M34 : implementation and second evaluation
M34-M36 : report preparation
Extra information
Prerequisite
Détails
Expected funding
Research contract
Status of funding
Expected
Candidates
Mathieu Golsenne
Utilisateur
nicolas.sabouret
Créé
Dimanche 05 janvier 2014 17:21:26 CET
dernière modif.
Dimanche 05 janvier 2014 17:21:26 CET
Fichiers joints
filename
créé
hits
filesize
Aucun fichier joint à cette fiche
Connexion
Ecole Doctorale Informatique Paris-Sud
Directrice
Nicole Bidoit Assistante
Stéphanie Druetta Conseiller aux thèses
Dominique Gouyou-Beauchamps
ED 427 - Université Paris-Sud
UFR Sciences Orsay
Bat 650 - aile nord - 417
Tel : 01 69 15 63 19
Fax : 01 69 15 63 87
courriel: ed-info à lri.fr