8th International Conference on Terminology and Artificial Intelligence - Toulouse (France),
November 18-20, 2009
Foreword
Rose
Dieng, former Director of Research at INRIA and organizer of TIA'07 in Nice,
was an active and dynamic member of the TIA group from its early days and a
driving force behind the continuation of the TIA conference series. Rose passed
away on June 30, 2008, and will be sadly missed. We would like to dedicate the
8th TIA conference in Toulouse to her memory.
Since
1995, the Terminology and Artificial
Intelligence (TIA) Conferences have focused on issues situated at the
crossroads of terminology and knowledge engineering. This has led to consider
problems related to knowledge acquisition from text and modelling of knowledge
on various information media and to reflect on the theoretical and
methodological consequences of the tasks these approaches imply as well as on
the applications derived from them. By focusing on corpora and the different
ways knowledge is expressed in running text, the TIA Conference quickly managed
to distinguish itself from other events that address similar topics.
In all specialized fields, knowledge is stored
and disseminated in the form of documents. In some cases, document management
(ranking according to relevance to a query, for instance) and acquisition and
extraction of knowledge for indexing, modelling and conceptualization rely on
terminological and linguistic studies. Hence, terminology and linguistics supply
key concepts to other disciplines such as natural language processing,
information science and knowledge engineering. Conversely, these disciplines
bring to light new problems and issues and present new challenges and
perspectives for terminology and linguistics. For nearly 15 years, TIA
conferences have provided a forum for researchers working in these various
fields to come together.
The TIA Conferences were first organized by a
research group with the same name (TIA group: http://tia.loria.fr/). The first edition took place in
Paris (TIA’95, Villetaneuse), and subsequent events have been held in a
variety of French cities (TIA’97 (Toulouse), TIA’99 (Nantes),
TIA’01 (Nancy), TIA’03 (Strasbourg), TIA’05 (Rouen), and TIA ’07
(Nice)). Created under the auspices of AFIA (l’Association
française pour l’intelligence artificielle; French Association for
Artificial Intelligence) – which has since become the AFIA/GdR-I3 group
– TIA brings together researchers in linguistics, natural language
processing and knowledge engineering. Although TIA Conferences have always
taken place in France, quickly, many researchers in other parts of the world
became interested in discussing the topics that were debated. The TIA group as
such no longer exists, but some of its instigators proposed that the conference
should remain and thus organized the 2009 edition.
This year, the programme committee wanted to
focus on terms, their realizations in running text and terminological
structures. Terms are used as focal points for knowledge structuring in many
applications, including ontologies, thesauri, and other conventional
terminological resources such as specialized dictionaries and terminological
databases. A number of semantic relations between terms may be identified (and
may vary depending on the application). However, although the aim of all of
these applications is a certain degree of stability, the linguistic nature of
terms and of the relations they share raises many questions. Which relations should
be represented in specific applications? How should terms and the relations
between them be represented? How can the relations in terminological resources
and their linguistic representation in texts be linked?
In line with what the founders of the Conference
aimed to create in 1995, TIA 2009 wants to provide an opportunity for
researchers developing fundamental and methodological approaches in knowledge
engineering in which textual linguistics plays a central role to discuss
problems and issues related to their work. A special emphasis has been placed
on multidisciplinary research in which linguistics, terminology, information
science and/or natural language processing are concerned. Hence, when selecting
articles, we paid special attention to the following criteria: diversity of
approaches, contribution against the background of previous work started more
than 10 years ago on the various links between text and knowledge
representation on formal media, studies and concrete experiments that help us
better understand the role of the term as a conveyor of specialized knowledge.
The programme committee first invited
contributions addressing theoretical Semantic theories and terminology in
relation to text linguistics and ontologies (especially theoretical linguistic
approaches to describing terms and terminological structures). By doing so, the
committee wanted to encourage participants to continue to investigate
fundamental issues.
In addition, the committee wanted to discuss
other methodological and theoretical questions such as: the representation of
terms and semantic and conceptual relations in specific data structures
(ontologies, thesauri, etc.), the representation of terms in relations in
multilingual applications, comparative studies of terminologies /
terminological resources / ontological resources from different languages,
communities or time frames; theoretical and technical problems in automated or
manual compilation of terminologies using mono- or multi-lingual corpora; methods
for automatic terminology structuring (identification of relations between
terms, linking of terms to specific fields of knowledge); studies on the
relationships between ontologies and terminologies and/or thesauri. Also,
submissions on complete or partial automation of knowledge acquisition and
representation were also invited, more specifically methods for structuring
terminologies automatically; identification of relations between terms,
automatic attribution of specialized domains, use of ontologies to better
understand their contribution to thesaurus or terminological semantics. Issues
related to evaluation methods and criteria, and validation of terminologies,
problems in compiling multilingual terminologies, reuse, standardization,
comparison and merging of terminological or ontological resources have also
always been identified as topics of interests at TIA Conferences. Finally,
applications are central in research work presented at TIA Conferences:
applications of terminological resources (the Semantic Web, information
retrieval, technology watch, question answering, document management, ranking
and/or classification, etc.).
The call for papers was a success since
32 articles were submitted to the programme committee. This year, the programme
has an international flavour since contributions were submitted by researchers
from Belgium, Canada, China, Croatia, France, Italy, Portugal, Spain, the
United Kingdom, and the United States. This also has a consequence on the
number of languages that are described in the papers which are much more varied
than in previous editions.
The programme committee selected 13 full papers
for oral presentation and 11 short papers presented as posters. Papers address
the following topics: semantic annotation of specialized and/or general corpora
for specific applications; ontology engineering or thesaurus compilation in
fields such as medicine, ceramics, the food industry, building construction,
environment; analysis of terms in corpora (e.g., polysemy, reduction of multiword
terms); special denominations, such as commercial trademarks; conceptual or
semantic patterns (this topic is addressed in a one of the two workshops
organized jointly with the Conference); images as media to convey knowledge;
applications such as automatic categorization of documents, compilation of rich
terminological resources (including formalized definitions or relations between
terms), term (or collocation) extraction.
In
addition to the papers selected by the programme committee, two invited
speakers have agreed to give talks that will highlight specific perspectives on
methods for acquiring knowledge from corpora and semantic analysis in running
text.
The
first talk is given by Diana Maynard and is entitled “GATE: Bridging the
Gap between Terminology and Linguistics”. The talk will describe the
basics of the GATE platform and of its components. It will also show that it
can lend itself to a variety of uses.
The second talk delivered by Carlos Subirats
and entitled “Spanish FrameNet. An On-line Lexical
Resource and its Application to NLP" will give a general overview of the Spanish version of FrameNet, how it
was compiled and how it can be applied to natural language processing.
This
year, two workshops were organized jointly with the Conference. In both cases,
the relation between text and knowledge is clearly emphasized. The first
workshop, entitled “Du thème au terme.
Émergence et lexicalisation des connaissances” (From Topic to
Term: Creation and Lexicalization of Knowledge), aim to better understand how
new concepts appear and evolve. The second workshop, entitled “Acquisition et modélisation de relations
sémantiques" (Acquiring and
Modelling Semantic Relations), focuses on fundamental aspects of the analysis
of semantic relations in specialized corpora, i.e. how to locate them in
running text and how to represent them in various types of applications.
The organization of TIA 2009
was made possible thanks to the support and help of many colleagues and friends
that we would like to thank here. First and foremost, we would like to thank
the organizers of the event, Nathalie Aussenac-Gilles and Anne Condamines,
who were very active in every aspect of the preparation of the Conference.
Their constant help and valuable advice made the organization of the conference
a much easier task to accomplish for us. The two Presidents of the organizing
committee could also count on a team comprising members from the IRIT (Institut
de Recherche en Informatique de Toulouse) and the CLLE-ERSS (Cognition, Langues,
Langage, Ergonomie-Equipe de Recherche en Syntaxe et Sémantique). For
the first time, two separate workshops are held jointly with the Conference and
we would like to thank their organizers – Sylvie Després, Natalia
Grabar, Monique Slodzian and Mathieu Valette – for posting original call
for papers on specific topics and for taking full responsibility of the
organization of the events.
We would also like to extend our thanks to the
members of the programme committee. Their advice and comments helped us design
a very rich programme. Also, we address special thanks to Diana Maynard and
Carlos Subirats for agreeing to give invited talks on topics of interest for
the participants.
Finally,
the organization of the two days of the Conference and of the day devoted to
the workshops could rely on financial contribution of the following French
organizations: the Conseil Régional de la Région Midi
Pyrénées, the Conseil National de la Recherche Scientifique
(CNRS), the Université Paul Sabatier Toulouse 3, the Université
Toulouse 2 le Mirail, the Institut National en Informatique et Automatique
(INRIA), and the Délégation Générale à la
Langue Française et aux Langues de France (DGLFLF), CFH: Facteurs
humains et ingénierie Linguistique. The Conference is held under the auspices of
the AFIA and the GDR I3.
Organizing committee
Under the responsibility of
Nathalie Aussenac-Gilles (IRIT, Toulouse) aussenac@irit.fr
Anne Condamines (CLLE-ERSS, Toulouse) anne.condamines@univ-tlse2.fr
Members
Françoise Agar (IRIT, Toulouse)
Jean-Pierre Bariteau (IRIT, Toulouse)
Ludovic Chacun (IRIT, Toulouse)
Jean-Philippe Cornille (IRIT, Toulouse)
Véronique Debats (IRIT, Toulouse)
Corinne Doumerc (IRIT, Toulouse)
Nathalie Duhaut (CLLE-ERSS, Toulouse)
Sabyne Lartigue (IRIT, Toulouse)
Stéphanie Lopez (CLLE - ERSS, Toulouse)
Christine Pernet (CLLE - ERSS, Toulouse)
Aurélie Piction (CLLE-ERSS, Toulouse)
Sophie Rességuier (IRIT, Toulouse)
Bernard Rothenburger (IRIT, Toulouse)
Anis Tissaoui (IRIT, Toulouse)
Programme
committee
Guadalupe Aguado (Universitad
Politécnica de Madrid, Espagne)
Amparo Alcina
(Universitat Jaume-I, Espagne)
Nathalie Aussenac-Gilles (IRIT, France)
Caroline Barrière (NRC, Canada)
Olivier Bodenreider (National Library of Medicine, USA)
Maria Teresa Cabré (Universitat Pompeu Fabra, Spain)
Marc van Campenhoudt (Université de Bruxelles, Belgique)
Farid Cerbah (Dassault-aviation, France)
Jean Charlet (AP-HP & INSERM, France)
Anne Condamines (CLLE-ERSS, France)
James Cussens (
Lyne Da Sylva (EBSI,
Montréal, Canada)
Valérie Delavigne (Institut national du cancer, France)
Patrick Drouin (OLST,
Montréal, Canada)
Pamela Faber (Universidad
de Granada, Espagne)
Ulrich Heid (Universität
Stuttgart, Allemagne)
Fidelia Ibekwe-SanJuan
(Université Lyon-3, France)
Kyo Kageura (University
of Tokyo, Japan)
Adeline Nazarenko
(LIPN, Université Paris 13)
Pascale Sébillot
(IRISA, France)
Koichi Takeuchi
(Okayama University, Japan)
Rita Temmerman
(Erasmushogeschool, Belgique)
Yannick Toussaint (LORIA,
France)
Pierre Zweigenbaum (LIMSI-CNRS & CRIM-INALCO, France)