15/06/2019
The Human Resources Strategy for Researchers

Large-scale oral data analysis and modeling for phonetic variation and dialogue adjustment

This job offer has expired


  • ORGANISATION/COMPANY
    CNRS
  • RESEARCH FIELD
    Language sciencesLanguages
    Language sciencesLinguistics
  • RESEARCHER PROFILE
    Recognised Researcher (R2)
  • APPLICATION DEADLINE
    06/07/2019 23:59 - Europe/Brussels
  • LOCATION
    France › ORSAY
  • TYPE OF CONTRACT
    Temporary
  • JOB STATUS
    Full-time
  • HOURS PER WEEK
    35
  • OFFER STARTING DATE
    01/10/2019

The work will consist in statistical exploration of large corpora that have been automatically pre-processed in order to model segmental variation marks such as reduction, assimilation or contextual dissimilation processes, voicing/devoicing patterns, marks of fluence / disfluence (pauses filled or empty, discourse markers) etc.

Speech varies in many ways and under the influence of various factors. This variation can be observed through differences in the pronunciations of words with respect to the so-called canonical reference, close to the norm encoded in dictionaries, and through many verbal events that are not words but contribute to the verbal message such as filled pauses (uh in French, uh and um in English etc.). Such marks are ubiquitous in the large speech corpora used in automatic speech processing and the investigation of their linguistic properties is relevant to speech technologies such as speech recognition (eg, the variation is encoded in pronunciation lexicons) or human – conversational agent dialogue (eg, the variation is modeled in order to manage and improve the naturalness of the verbal interaction).
We propose an interdisciplinary topic, involving phonetics and phonology, statistical modeling of linguistic data and automatic speech processing (speech recognition, human-conversational agent dialogue). The topic deals with the analysis of variation in large corpora in French and English in relation to phonetic adjustment phenomena specific to verbal interaction. The aim is to estimate the interaction between variation phenomena continuous speech proper (eg, segmental reduction) and dialogue specific phonetic adjustment, and to contribute to pattern modeling in order to improve human-conversational agent communication.

This postdocotral work is conducted within a joint project between LIMSI CNRS (Ioana Vasilescu, Alexandre Allauzen) and Télécom ParisTech (Chloé Clavel) supported by the Labex Digicosme. The candidate will work within the “Speech Processing Group” at LIMSI CNRS.

Eligibility criteria

-Phonetics, phonology, large scale corpora, statistical analysis, discourse analysis
-Praat, Rstudio,Perl.

Additional comments

NA.

Web site for additional job details

Required Research Experiences

  • RESEARCH FIELD
    Language sciences
  • YEARS OF RESEARCH EXPERIENCE
    1 - 4
  • RESEARCH FIELD
    Language sciencesLinguistics
  • YEARS OF RESEARCH EXPERIENCE
    1 - 4
  • RESEARCH FIELD
    Language sciencesLanguages
  • YEARS OF RESEARCH EXPERIENCE
    1 - 4

Offer Requirements

  • REQUIRED EDUCATION LEVEL
    Language sciences: PhD or equivalent
  • REQUIRED LANGUAGES
    FRENCH: Basic
Work location(s)
1 position(s) available at
Laboratoire d'Informatique pour la Mécanique et les Sciences de l'Ingénieur
France
ORSAY

EURAXESS offer ID: 417378
Posting organisation offer ID: 10036

Disclaimer:

The responsibility for the jobs published on this website, including the job description, lies entirely with the publishing institutions. The application is handled uniquely by the employer, who is also fully responsible for the recruitment and selection processes.

 

Please contact support@euraxess.org if you wish to download all jobs in XML.