LINGUIST List 14.3491

Tue Dec 16 2003

Calls: Ling & Literature/Canada; Text/Corpus Ling

Editor for this issue: Andrea Berez <andrealinguistlist.org>

Directory

1. gacerenz, Dictionnaires Fran�ais et Litt�ratures Qu�b�coise et Canadienne-fran�aise
2. thomas.beauvisage, Workshop on Characterisation of Internet Content

Dictionnaires Fran�ais et Litt�ratures Qu�b�coise et
Canadienne-fran�aise

Date: 11-Nov-2004 - 13-Nov-2004
Location: Waterloo (Ontario), Canada
Contact: Gerardo Acerenza
Contact Email: gacerenzwatarts.uwaterloo.ca 
Meeting URL: http://french.uwaterloo.ca 

Linguistic Sub-field: Ling & Literature 
Subject Language: French 
Call Deadline: 30-Mar-2004 

Meeting Description:

Ce colloque se propose d'examiner le r�le que les dictionnaires
fran�ais jouent dans les litt�ratures francophones d'Am�rique. Il
s'agit de comprendre comment les dictionnaires venant de France
caract�risent, enrichissent ou mod�lent les productions litt�raires
qu�b�coises et canadiennes-fran�aises. Il importe aussi de voir quelle
est la place que les litt�ratures qu�b�coise et canadienne-fran�aise
occupent dans les dictionnaires de France. 

APPEL DE COMMUNICATIONS

DICTIONNAIRES FRANCAIS ET LITTERATURES QUEBECOISE ET
CANADIENNE-FRANCAISE

Colloque organis� par le D�partement d'�tudes fran�aises et italiennes
de l'Universit� St. J�r�me en collaboration avec le D�partement
d'�tudes fran�aises de l'Universit� de Waterloo Waterloo, Ontario,
Canada, 11-13 novembre 2004

Ce colloque se propose d'examiner le r�le que les dictionnaires
fran�ais jouent dans les litt�ratures francophones d'Am�rique. Il
s'agit de comprendre comment les dictionnaires venant de France
caract�risent, enrichissent ou mod�lent les productions litt�raires
qu�b�coises et canadiennes-fran�aises. Il importe aussi de voir quelle
est la place que les litt�ratures qu�b�coise et canadienne-fran�aise
occupent dans les dictionnaires de France. Au Qu�bec et au Canada
fran�ais, les �crivains consultent, lisent et parfois �tudient avec
acharnement les dictionnaires venant de France (Robert, Larousse,
Littr�, etc.). Pour les uns, cet immense r�servoir de mots qu'est le
dictionnaire repr�sente une bou�e de sauvetage. Pour les autres, c'est
une autorit� normative qui ne reconna�t pas toutes les vari�t�s
langagi�res de la francophonie. �La langue fran�aise est tr�s
difficile � contr�ler. C'est mon outil de travail. Je v�rifie souvent
dans le dictionnaire� avoue Francine No�l. Parfois, le dictionnaire
devient source de cr�ation litt�raire et artistique. Que l'on pense au
Terrain du dictionnaire A/Z de l'�crivain qu�b�cois Rober Racine. Il
d�coupe patiemment les 55.000 mots et d�finitions du Robert pour les
coller ensuite sur des b�tonnets noirs et les placer debout sur une
surface blanche, dans le but de �pr�senter autrement [ce
dictionnaire]; le rendre visible� � un �lecteur promeneur� qui se
d�place physiquement d'un mot � un autre (Le dictionnaire). Pour Rober
Racine, le dictionnaire est avant tout un objet de fascination, voire
une source d'inspiration. Pour un grand nombre d'�crivains, il
repr�sente la r�f�rence, la norme. Jacques Ferron, par exemple,
semblait �tre un fervent admirateur de Littr� et du Littr� qu'il
�voquait chaque fois avec d�f�rence: �Je ne me sers que de Littr�. '
Je n'ai jamais invent� de mots.' Je suis � ce sujet de stricte
observance� (Du fond de mon arri�re-cuisine). Les communications, en
fran�ais ou en anglais, d'une longueur de 20 minutes, suivront ces
trois axes de r�flexion:
1) Les �crivains qu�b�cois et canadiens-fran�ais et les dictionnaires:
t�moignages � partir d'autobiographies, journaux intimes,
correspondances ou m�moires. Que repr�sentent les dictionnaires
fran�ais pour les �crivains qu�b�cois, acadiens ou franco-ontariens
des XIXe et XXe si�cles? Y a-t-il des fanatiques du/des
dictionnaire/s? Quel usage font-ils des dictionnaires? En font-ils un
usage standard? Manifestent-ils leur accord ou leur d�saccord
vis-�-vis des dictionnaires de France?
2) Les oeuvres qu�b�coises et canadiennes-fran�aises et les
dictionnaires: roman, po�sie, th��tre. Les textes qu�b�cois, acadiens
ou franco-ontariens pr�sentent-ils des marques explicites de la
pr�sence des dictionnaires fran�ais? De quelle mani�re les
dictionnaires enrichissent-ils, caract�risent-ils ou mod�lent-ils les
oeuvres? Dans quels contextes fictionnels les dictionnaires sont-ils
�voqu�s?
3) Les dictionnaires et les litt�ratures qu�b�coise et
canadienne-fran�aise: la pr�sence des "uvres litt�raires dans les
dictionnaires. Quels sont les �crivains qu�b�cois et
canadiens-fran�ais cit�s dans les dictionnaires de France? Sont-ils
cit�s dans le but de l�gitimer des usages langagiers particuliers?
Contribuent-ils � enrichir la langue fran�aise ou bien sont-ils cit�s
comme exemples de style?

Pri�re d'envoyer les propositions de communication (250-300 mots)
avant le 30 mars 2004 aux responsables du colloque:

Gerardo Acerenza (519) 884-8111, poste 227. Courriel:
gacerenzwatarts.uwaterloo.ca (fran�ais)

Gabriel Niccoli (519) 884-8111, poste 235. Courriel:
gniccoliuwaterloo.ca (anglais)

Workshop on Characterisation of Internet Content 

Date: 31-Jan-2004 - 31-Jan-2004
Location: Paris, France
Contact: Beauvisage Thomas
Contact Email: indices.internetml.free.fr 
Meeting URL: http://www.atala.org/je 

Linguistic Sub-field: Text/Corpus Linguistics, Semantics,
Computational Linguistics

Call Deadline: 05-Jan-2004 

Meeting Description:

This workshop is organised to point to existing problems faced by NLP
tools for the description and use of material available on the
Internet (Web pages and sites, mail, fora, instant mail, etc). Such
problems concerns in particular:
	- methods for content collection, accessibility of content,
formalisms for information storage, etc.
 - the semantics of Internet content: textual content
vs. services offered, multimedia and interactive content, semiotics of
web pages;
- categorisation methods: topic detection, categorisation of sites and
pages, Internet-specific genres. Globally speaking, these problems
reveal new links between NLP and the Internet. On the same time,
because of these problems, computational linguistics has to study
further the benefits and limitations of NLP tools for the description
and use of Internet content. 

Workshops of the Association pour le Traitement Automatique des
Langues (ATALA)

CALL FOR SUBMISSIONS

Location:
ENST, 46, rue Barrault, 75013 Paris, M�tro : Corvisart

Date:
SATURDAY JANUARY 31TH, 2004

Subject:
CHARACTERISATION OF INTERNET CONTENT: BEYOND KEYWORDS, SEMANTIC
APPROACH.

Workshop organised by Fran�ois Rastier (CNRS - UMR 7114, Paris X -
MoDyCo), Natalia Grabar (CRIM/INaLCO, STIM / DSI / AP-HP, Paris 6) and
Thomas Beauvisage (France T�l�com R&D - DIH/UCE, Paris X - MoDyCo)

[This call for submissions is also available on the ATALA website (www.atala.org)]


Issues:
Applications related to characterisation, detection and processing of
Internet content (Web, mail, fora, ...) are ever more numerous, both
in relation to Internet access and content filtering. Therefore, there
exists a real social need which is direclty related to NLP techniques:
NLP appears here as the user of Internet content, but also as the
developer of tools for the content processing. But while all documents
available on the Web and network exchanges (mails, web discussion
fora, ...) make up a digitised text repository that is constantly
expanding, their detection, collection and processing raise a series
of technical and theoretical problems.

First of all, NLP tools have to be adapted to the lexical content of
Internet documents (specific vocabulary, formatting, check spelling,
grammatical correction). More generally, the poly-semiotic nature of
Web content raises questions about the relevance of lexicon-specific
content processing. Two examples can help us realise the usefulness of
going beyond lexical approaches:
- search engines have significantly improved their performances with
the exploitation of structural elements of HTML pages (''keywords''
tags, ...) and Web links.
- in filtering applications, key-words seem to be insufficient. They
make up a baseline and have to be combined with other types of
features (morphology, punctuation, syntax, etc.). Pictures,
formularies, layout, evolutivity, services offered, exchange
structures are some other elements which require considering Internet
content as a specific activity, with its own interaction modes and its
specific user-centered rules.

Because of these problems, computational linguistics has to study
further the use of Internet content. It is then important to define
which textual and non-textual elements -going beyond merely lexical
material- participate in the characterisation of Internet content and
which kinds of tools are useful to bring this characterisation to
light. This workshop is organised to point to existing problems and
current solutions in different applications.

Objectives
This workshop is organised to point to existing problems faced by NLP
tools for the description and use of material available on the
Internet (Web pages and sites, mail, fora, instant mail, etc). Such
problems concerns in particular:
	- methods for content collection, accessibility of content,
formalisms for information storage, etc.
	- the semantics of Internet content: textual content
vs. services offered, multimedia and interactive content, semiotics of
web pages;
	- categorisation methods: topic detection, categorisation of
sites and pages, Internet-specific genres. Globally speaking, these
problems reveal new links between NLP and the Internet. On the same
time, because of these problems, computational linguistics has to
study further the benefits and limitations of NLP tools for the
description and use of Internet content.

We are particularly interested in work which goes beyond the
single-criterion analysis (i.e. keywords) and offers an analysis which
takes into account different organisational levels of the document
 - inside the ergonomic unity of the document: textual
(lexical, grammatical, etc.), visual (pictures, logos), structural
(text, peritext) or other elements (frames),
- the environment of the document: the sites or the services
proposed within the document, the network of pages which the document
is part of (internal, external links, anchors), the usage scenarii in
which the document exists, etc.
- in the intertextual integration of the document on the
Internet (connectivity, hypertext).

Submission
Interested authors can send a 2-to-4-page abstract of their work,
providing the following information:

- research objectives and application fields,
- stage of research work,
- theoretical and practical research-related issues,
- bibliography.


Abstracts have to be sent to the following e-mail address: 
indices.internetml.free.fr
(this adress will be close on Feb. 1st 2004.

Accepted document formats (in preference order): PDF, PS, TXT, DOC,
RTF

Official languages: French, English

Important dates
Submission: January 05, 2004 
Notification: January 10, 2004 
Workshop: January 31, 2004