Seminaire: Serie de seminaires, A. Villavicencio, labex Transfers, Paris et Montrouge

Wed Apr 23 19:28:26 UTC 2014

Date: Wed, 23 Apr 2014 15:53:54 +0200
From: Thierry Poibeau <thierry.poibeau at ens.fr>
Message-Id: <99BB7B0C-1587-42DA-8CD7-156D14F2C355 at ens.fr>
X-url: http://transfers.ens.fr/

Bonjour, 

Veuillez trouver ci-joint le programme des conférences d’Aline
Villavicencio (http://www.inf.ufrgs.br/~avillavicencio/), invitée du
labex Transfers (http://transfers.ens.fr/) en mai 2014.

L’ensemble des conférences sera consacré à l’acquisition du langage et à
la modélisation (notamment informatique) des phénomènes correspondants.

Toutes les conférences seront en anglais. 

Cordialement,

Thierry Poibeau

====================

Short Bio:
Aline Villavicencio is a senior lecturer in Computer Sciences at the
Federal University of Rio Grande do Sul (Brazil), and a CNPq fellow. She
was a visiting Scholar at MIT (USA) from 2011-2012, and at Saarland
University in 2012-2013, with PhD and MPhil degrees from the University
of Cambridge, UK. Her research has included work on computational
language acquisition and grammar engineering for languages such as
English and Portuguese. She has coordinated several projects on these
topics, which include collaboration with France, US and Latin American
universities. She has organized events including the ACL-2007 and the
EACL-2009, 2012 and 2014 Workshop on Cognitive Aspects of Computational
Language Acquisition, and the ACL 2003, 2004, 2011, NAACL-2013 and
Coling 2010 workshops on Multiword Expressions among others.     

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

Topic: Language Acquisition
Title: Language Acquisition and Probabilistic Models: keeping it simple
Joint work with Bob Berwick (MIT)

Abstract: Hierarchical Bayesian Models (HBMs) have been used with some
success to capture empirically observed patterns of under- and
overgeneralization in child language acquisition. However, as is well
known, HBMs are ``ideal'' learning systems, assuming access to unlimited
computational resources that may not be available to child language
learners. Consequently, it remains crucial to carefully assess the use
of HBMs along with alternative, possibly simpler, candidate models. In
this talk I describe an evaluation for a language acquisition domain
where explicit HBMs have been proposed: the acquisition of English
dative constructions. In particular, I present a detailed,
empirically-grounded model-selection comparison of HBMs vs. a simpler
alternative based on clustering along with maximum likelihood
estimation, the linear competition learning (LCL). Results obtained
demonstrate that LCL can match HBM model performance without incurring
on the high computational costs associated with HBMs.

Mercredi 7 mai
10h30 - 12h30, ENS 45 rue d’Ulm salle Cartan

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

Topic: Language Acquisition work with psycholinguistic data
Title: Lexical and topological analysis of language acquisition and
dissolution 

Abstract: Psycholinguistic studies on language acquisition, organization
and dissolution have proposed several hypotheses to explain human
processing of language, where factors such as frequency, polysemy and
conventionality seem to influence language use. Recent studies have also
indicated the impact of gender and age, with differences in language use
between boys and girls found in longitudinal data. In this talk I will
describe lexical and topological analyses of some target groups,
including children for language acquisition, and clinical groups for
language dissolution investigations.

Mardi 20 mai
10h00 - 12h30, Laboratoire Lattice (salle 512, 1 rue Maurice Arnoux
92120 Montrouge)

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

Topic: Multi Word Expressions
* Title: Multiword expressions in language technology and language
acquisition

Abstract: Conventional ways of transmitting efficiently and concisely
complex ideas and concepts in natural language in a particular domain or
community often adopt Multiword Expressions (MWE). For example, "an old
wives' tale" as "an idea believed in the past but now known to be
wrong".  MWEs are a heterogeneous set that includes various complex
phenomena ranging from compound nouns and verb particle constructions to
idioms, and have received considerable attention from researchers from a
variety of disciplines, such as Linguistics, Psycholinguistics, and
Natural Language Processing. Although their importance has long been
recognised, their heterogeneous characteristics have proved a challenge
to both linguistic and computational analysis. In this talk I examine
some of these challenges and present an approach for their automatic
identification. I then discuss some applications for their use in
language technology and child language acquisition studies.

suivi de :

------MWEs methods
* Title: A language independent framework for Multiword Expressions
Identification

Abstract: A Multiword Expression (MWE) is a sequences whose
interpretation crosses the word boundaries (Sag et al., 2002) and
include compound nouns (credit card, mountain bike), phrasal verbs
(carry on, go by [a name]), compound terms (benign tumor, nuclear
fusion), etc.  They are an integral part of language, and their
importance has long been recognized. However, their heterogeneous
characteristics have proved a challenge to both linguistic and
computational analysis. In this paper we discuss some of these
challenges along with a variety of approaches for semi-automatically
finding and handling these expressions.

Mercredi 21 mai
10h00 - 12h30, Laboratoire Lattice (salle 512, 1 rue Maurice Arnoux
92120 Montrouge)

-------------------------------------------------------------------------
Message diffuse par la liste Langage Naturel <LN at cines.fr>
Informations, abonnement : http://www.atala.org/article.php3?id_article=48
English version       : 
Archives                 : http://listserv.linguistlist.org/archives/ln.html
                                http://liste.cines.fr/info/ln

La liste LN est parrainee par l'ATALA (Association pour le Traitement
Automatique des Langues)
Information et adhesion  : http://www.atala.org/

ATALA décline toute responsabilité concernant le contenu des
messages diffusés sur la liste LN
-------------------------------------------------------------------------