28.2545, Support: French; Old French; Computational Linguistics: PhD, LaTTiCe, Université Sorbonne Nouvelle

[The LINGUIST List]

LINGUIST List: Vol-28-2545. Thu Jun 08 2017. ISSN: 1069 - 4875.

Subject: 28.2545, Support: French; Old French; Computational Linguistics: PhD, LaTTiCe, Université Sorbonne Nouvelle

Moderators: linguist at linguistlist.org (Damir Cavar, Malgorzata E. Cavar)
Reviews: reviews at linguistlist.org (Helen Aristar-Dry, Robert Coté,
                                   Michael Czerniakowski)
Homepage: http://linguistlist.org

Please support the LL editors and operation with a donation at:
           http://funddrive.linguistlist.org/donate/

Editor for this issue: Clare Harshey <clare at linguistlist.org>
================================================================


Date: Thu, 08 Jun 2017 14:03:22
From: Isabelle Tellier [isabelle.tellier at univ-paris3.fr]
Subject: French; Old French; Computational Linguistics: PhD, LaTTiCe, Université Sorbonne Nouvelle, Paris, France

 Institution/Organization: Lattice 
Department: CNRS, ENS, université Sorbonne Nouvelle - Paris 3 
Web Address: http://www.lattice.cnrs.fr/ 

Level: PhD 

Duties: Research
 
Specialty Areas: Computational Linguistics 
 
Required Language(s): French (fra)
                      French, Old (fro) 

Description:

3 year funded PhD position: annotation and syntactic analysis of heterogeneous
corpora
The ANR funded Profiterole project (PRocessing Old French Instrumented TExts
for the Representation Of Language Evolution, see
http://www.agence-nationale-recherche.fr/?Projet=ANR-16-CE38-0010) is seeking
a candidate for a 3-year doctoral position.

One of the main features of medieval French is its great variability at the
phonetic, morphological as well as syntactical levels: the thesis will be
devoted to the exploration of heterogeneous (especially medieval) corpora
using machine learning techniques and formal grammars.

On the machine learning side, the emphasis will be on adapting the labeled
corpora used as training sets. The purpose will be to define a general
methodology for acquiring the best tagger (parts of speech) and the best
parser (syntactic functions) for a given new text, making the best use of
already annotated texts.

On the formal grammar side, the adaptation of meta-grammars defined for modern
French will be explored.

The methodology will include:
1. Standardization procedures to be applied to texts;
2. Exploration of already annotated texts to identify their most
discriminative properties. The objective will be to design a decision
procedure able to identify, for a new target text, which already annotated
texts are most likely to produce an adapted model.
3. Adaptation to medieval French of meta-grammar defined for modern French;
4. Development of strategies for the automatic detection of annotation errors,
their quantitative and qualitative analysis;
5. Combination of different parsers with ensemble techniques.

NLP programs are more and more confronted with heterogeneous linguistic data.
Medieval texts are a case study of particular relevance to address this
problem, but the methodology developed in this thesis will be general enough
to have applications far beyond this specific context.

Requirements:
Position requirements include a Master degree in NLP, with a solid background
in Machine Learning and familiarity with syntactic formalisms. Knowledge in
Medieval French would be an advantage though not essential.

Duration: 36 months, starting from September 2017.
Salary: the net salary per month amounts to 1536 euros

Location (Paris): 
Lattice laboratory,
Ecole Normale Supérieure
1 rue Maurice Arnoux
F-92120 Montrouge

Applications should include a letter of application, a curriculum vitae and
academic marks to be submitted to the email addresses below by June 30. 
 

Application Deadline: 30-Jun-2017 

Mailing Address for Applications:
	Attn: Sophie Prévost 
	Lattice, 1 rue Maurice Arnoux 
	Montrouge 92120 
	France
	
	sophie.prevost at ens.fr 
	isabelle.tellier at sorbonne-nouvelle.fr
	Eric.De_La_Clergerie at inria.fr 
	
Web Address for Applications: http://www.lattice.cnrs.fr/ 

Contact Information: 
	PR Isabelle Tellier 
	isabelle.tellier at sorbonne-nouvelle.fr 
	Phone:688233355  


------------------------------------------------------------------------------

*****************    LINGUIST List Support    *****************
Please support the LL editors and operation with a donation at:
            http://funddrive.linguistlist.org/donate/
 


----------------------------------------------------------
LINGUIST List: Vol-28-2545	
----------------------------------------------------------