COMPUTATIONAL APPROACHES TO DISCOURSE AND DOCUMENT PROCESSING
 
*** Special Issue of the Journal "Traitment Automatique des Langues"
    (TAL) ***
 
Deadline for submission: 1st September 2006
 
 
GUEST EDITORS: Marie-Paule Piry-Woodley (ERSS - Universiti de Toulouse
2, France) and Donia Scott (Open University, UK)
 
FIRST CALL FOR PAPERS
 
This special issue of TAL is devoted to what we perceive as an
increasing convergence between discourse linguistics, document
engineering and NLP. Such a convergence can be observed from several
angles; for example, new modes of access to documents' contents place
greater emphasis on exploiting discourse structure; applying corpus
analysis methods to discourse calls for greater use of NLP
techniques.  It is manifest in the increasing number of joint studies,
and results in cross fertilisation of the disciplines.
 
*Text and discourse linguistics
Constructing a coherent interpretation of discourse involves
delimiting segments, and identifying semantico-pragmatic relations and
structures that bind them. These processes may be approached via the
notions of discourse relations (cf. RST, SDRT), of theme or topic, of
discourse framing, etc. A major challenge for the field is to identify
linguistic correlates of specific discourse functions in text corpora,
taking into account potential register- and domain-linked
variation. Another promising research area focuses on the global
structure of documents: role of titles, "logical structure", or
layout, to name a few.
 
* Document engineering and NLP
Access to the information stored in electronic documents is a major
issue, and recent approaches seek to take better account of the
organisation of the documents being processed - such as their thematic
and rhetorical structure. A growing number of applications are
concerned by this evolution: automatic summarisation, document
browsing, information retrieval (e.g. for passage extraction from
selected documents). In addition, hypertextual and composite documents
raise new questions about the interaction between semiotic functions
(text-image for instance).
 
For this special issue of TAL, we invite papers from researchers
working in the fields of discourse linguistics, computational
linguistics and document engineering on what can be termed the
"document level", whereby:
 
-         the document is envisaged as a functional unit, situated in
          a specific setting,
and actualised in a medium (whatever its nature) which confers on it
          material characteristics which are an integral part of the
          use and sense made of it;
 
- the expression "document level" signals a marked interest in coarse
  grain structures, or "global" structures (so-called "logical"
  structure, document structure, organisation in sections), or in the
  interaction between local and global structures (reference chains,
  topical chains, inter-propositional relations, etc.). It indicates a
  specific interest in "top down" approaches based on the exploitation
  of surface cues or markers.
 
Specific topics include (non-exhaustive list):
- Analysis of discourse structures;
- Analysis of the structure of composite documents (text-images,
text-graphics);
- Impact of hypertextual or hyperdocument organisation on the way
documents are produced and understood. Hypertext formatting,
computational tools based on NLP procedures;
- Reading models, discourse organisation models;
- Computational experimentation as a means of testing hypotheses on
corpora;
- Preliminary corpus studies: identification and annotation of
discourse structures;
- Computational devices for such experimental approaches (formalising
linguistic knowledge, ensuring re-usability of resources, articulating
levels of processing, visualising annotations);
- Articulating approaches using linguistic markers, domain knowledge,
statistical techniques;
- Applications: intra-document information retrieval, browsing aids,
document summarisation or synthesis, etc.
 
THE JOURNAL
 
(see http://www.atala.org/)
TAL (Traitement Automatique des Langues/Natural Language Processing)
is a forty year old international journal edited by ATALA (French
Association for Natural Language Processing) with the support of CNRS
(National Centre for Scientific Research).
It is now moving over to an electronic mode of publication, with
printing on demand. This will in no way affect the reviewing and
selection process.
 
LANGUAGE
 
Manuscripts may be submitted in English or French. French-speaking
authors are requested to submit in French.
 
IMPORTANT DATES
 
01/09/2006 Deadline for submission
23/10/2006 Notification to authors
30/11/2006 Deadline for submission of revised version
 
PAPER SUBMISSION
 
Contributions (25 pages maximum, PDF format) will be sent by e-mail to
both editors at the addresses below:
Marie-Paule Piry-Woodley <pery@univ-tlse2.fr>
Donia Scott < D.Scott@open.ac.uk>
 
Style sheets are available for download at
http://tal.e-revues.com/appel.jsp
 
SPECIAL EDITORIAL BOARD
 
N. Asher (IRIT, U. Toulouse 3, France)
J. Bateman (U. Bremen, Germany)
Y. Bestgen (U. C. Louvain, Belgium)
N. Bouayad-Agha (U. Pompeu Fabra, Barcelona, Spain)
M. Charolles (U. Paris 3, France)
D. Cristea (U. Iasi, Romania)
L. Danlos (U. Paris 7, France)
L. Degand (U. C. Louvain, Belgium)
P. Enjalbert (U. Caen, France)
S. Ferrari (U. Caen, France)
B. Grau (U. Paris-Sud, France)
N. Hernandez (GREYC, U. Caen, France)
J. Karlgren (Swedish Institute of Computer Science, Sweden)
G. Lapalme (U. Montrial, Quibec, Canada)
N. Lucas (GREYC, U. Caen, France)
A. Max (U. Paris-Sud, France)
J.-L. Minel (U. Paris 4, France)
R. Power (Open University, England)
H. Saggion (U. Sheffield, England)
S. Teufel (U. Cambridge, England)