By continuing to browse the site, you are agreeing to our use of cookies. (More details)

Nudge detection in Human-Machine interaction and Human-human Conversation (H/F)

This offer is available in the following languages:
Français - Anglais

Ensure that your candidate profile is correct before applying. Your profile information will be added to the details for each application. In order to increase your visibility on our Careers Portal and allow employers to see your candidate profile, you can upload your CV to our CV library in one click!

Faites connaître cette offre !

General information

Reference : UPR3251-LAUDEV-002
Workplace : ORSAY
Date of publication : Thursday, June 25, 2020
Scientific Responsible name : Laurence Devillers
Type of Contract : PhD Student contract / Thesis offer
Contract Period : 36 months
Start date of the thesis : 15 September 2020
Proportion of work : Full time
Remuneration : 2 135,00 € gross monthly

Description of the thesis topic

Nudges, indirect suggestions which can affect the behaviour and the decision making, are considered in the context of conversational machines.The notion of “nudging” first came into light in 2008 when it was pro-posed by R.H. Thaler (Nobel Prize in Behavioral Economy, Nov 2017)and C.R. Sunstein. They stressed the fact that “nudging” was a tactic to subtly modify a person's behavior, without restricting that person's choice. Indeed, nudging mainly operates through the affective system orby exploiting common cognitive bias (e.g. attention, memory, lazyness).The thesis will focus on Nudge detection in spoken dialogs from Human-Machine interaction and Human-Human Conversation. Audio mimicry(such as facial mimicry) is a mark of engagement between two persons in social interaction. Jointly decoding lexical and audio channels could help to disambiguate semantic content but also to compute a metric of alignment between 2 persons by making use of the signal within disfluencies and other audio features that characterize the most human form of information transmission and generation: conversational communication.In particular, leveraging the audio channel in the segmentation of spoken conversation could produce more flexible, robust, and meaningful units. Among several challenging facts, they raise the issue of the ethical use with respect to sensitive populations such as children or elderly who are not sufficiently aware of the human technology and intentions behind such devices. We have already collected corpora with nudges strategies in the dialog involving primary school children and students at LIMSI-CNRS which could be used for this study.

Work Context

This thesis is financed within the con-text of the Chair AI HUMAAINE: HUman MAchine Affective InteractionEthics at LIMSI-CNRS. The LIMSI (Laboratoire d'informatique pour la mécanique et les sciences de l'ingénieur) is a research unit of the "Centre national de la recherche scientifique" (CNRS). Associated with the University of Paris-Sud, this laboratory conducts research on two major themes: mechanics and energy; man-machine communication.

We talk about it on Twitter!