Faites connaître cette offre !
Reference : UMR5267-AGAJAC-001
Workplace : MONTPELLIER
Date of publication : Tuesday, December 11, 2018
Type of Contract : FTC Scientist
Contract Period : 18 months
Expected date of employment : 1 February 2019
Proportion of work : Full time
Remuneration : -about 1950 € (net)
Desired level of education : PhD
Experience required : Indifferent
The successful applicant will be given the mission to define and structure, by means of an ontology, the linguistic knowledge related to reference and nomination.
Working with all of the participants of the TALAD project, the post-doc will devote an in-depth reflection to (i) the scientific, methodological and technological interaction between two communities: discourse analysis (DA/AD) and natural language processing (NLP/TAL), (ii) the proper matching between an operational semantics of nomination and the actual needs in policy and societal intelligence.
The post-doc will be responsible for the design of corpora annotation schemes with the concepts of the ontology and will monitor the production task of annotated corpora. To this end, in parallel with the manual annotation, new semi-automatic annotation methods based on active learning might be explored.
He/she will also responsible for the validation and distribution of all models and resources, including methodological guidelines for the analysis practices that will be defined in the project.
Particular attention will be paid to the links possibly established between denomination and two other phenomena partially operative in NLP: named entities and coreferences.
Have completed a Ph.D. degree in computational linguistics, machine learning, computer science or a related discipline.
Master the techniques used in natural language processing and/or standards for knowledge modeling with ontologies (OWL, SKOS, Lexical Markup Framework - LMF, Lexicon Model for Ontologies - LEMON).
This post-doctoral position of 18 months is funded by the TALAD « Analyse et traitement automatique de discours » ANR project involving AGORA EA7392 (Université Cergy Pontoise), ERTIM (Inalco), LI (Université de Tours), PRAXILING (Université Paul-Valéry) and the company RETICULAR PROJECT (Paris).
At the interface between DA and NLP, the TALAD project deals with theorizing and modeling the complex language dynamics used in the reference act. Work is planned in particular on automatic recognition of various nomination/denomination/designation patterns, supporting opinions or standpoints in texts.
For a complete presentation of the project, see:
The application file includes a résumé, copy of the degree, a scientific project, a letter of motivation and possibly recommendation letters.
We talk about it on Twitter!