[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[eu_members at aclweb dot org] Singapore: ACL/IJCNLP 2009 2nd Workshop on Building and Using Comparable Corpora -- EXTENDED DEADLINE



=======================================================================
       Extended call for papers
     New deadline for submission: May 15th, 2009

2nd Workshop on Building and Using Comparable Corpora:
from parallel to non-parallel corpora

   ACL-IJCNLP 2009
=======================================================================

================================================
August 6th, 2009
Suntec, Singapore
http://comparable2009.ust.hk/

Submission:
https://www.softconf.com/acl-ijcnlp09/BUCC/
  short papers: 4 pages
  long papers: 8+1 pages
================================================


OBJECTIVE

 Following the success of the first Workshop on Building and Using
 Comparable Corpora
 <http://www.limsi.fr/~pz/lrec2008-comparable-corpora/> at LREC 2008,
 this workshop aims to bring together language engineers as well as
 linguists interested in the constitution and use of comparable
 corpora, ranging from parallel to non-parallel corpora. In the larger
 context of the joint ACL-IJCNLP, this workshop aims to solicit
 contributions from researchers in different geographical regions, in
 order to highlight in particular the issues with comparable corpora
 across languages that are very different from each other, such as
 across Asian and European languages. Research in minority languages
 is also of particular interest.


MOTIVATION

 Research in comparable corpora has been motivated by two main reasons
 in the language engineering and the linguistics communities. In
 language engineering, it is chiefly motivated by the need to use
 comparable corpora as training data for statistical NLP applications
 such as statistical machine translation or cross-lingual retrieval.
 In linguistics, on the other hand, comparable corpora are of interest
 themselves in providing intra-linguistic discoveries and
 comparisons. It is generally accepted in both communities that
 comparable corpora are documents in one to many languages, that are
 comparable in content and form in various degrees and dimensions. It
 was pointed out that parallel corpora are at one end of the spectrum
 of comparability whereas quasi-comparable corpora are at the other
 end. We believe that the linguistic definitions and observations in
 comparable corpora can improve methods to mine such corpora for
 applications to statistical NLP. As such, it is of great interest to
 bring together builders and users of such corpora.

 Parallel corpora are a key resource as training data for statistical
 machine translation, and for building or extending bilingual lexicons
 and terminologies. However, beyond a few language pairs such as
 English-French or English-Chinese and a few contexts such as
 parliamentary debates or legal texts, they remain a scarce resource,
 despite the creation of automated methods to collect parallel corpora
 from the Web. Interests in non-parallel forms of comparable corpora
 in language engineering primarily ensued from the scarcity of
 parallel corpora. This has motivated research into the use of
 comparable corpora: pairs of monolingual corpora selected according
 to the same set of criteria, but in different languages or language
 varieties. Non-parallel yet comparable corpora overcome the two
 limitations of parallel corpora, since sources for original,
 monolingual texts are much more abundant than translated
 texts. However, because of their nature, mining translations in
 comparable corpora is much more challenging than in parallel
 corpora. What constitutes a good comparable corpus, for a given task
 or per se, also requires specific attention: while the definition of
 a parallel corpus is fairly straightforward, building a non-parallel
 corpus requires control over the selection of source texts in both
 languages.

 With the advent of online data, the potential for building and
 exploring comparable corpora is growing exponentially. Comparable
 documents in languages that are very different from each other pose
 special challenges as very often, the non-parallel-ness in sentences
 can result from cultural and political differences.


INVITED SPEAKER

 Kenneth Ward Church (Microsoft Research, Redmond)


TOPICS

 We solicit contributions in but not limited to the following topics:

 * Building Comparable Corpora
     - Human translations
     - Automatic and semi-automatic methods
     - Methods to mine parallel and non-parallel corpora from the Web
     - Tools and criteria to evaluate the comparability of corpora
     - Parallel vs non-parallel corpora, monolingual corpora
     - Rare and minority languages
     - Across language families
     - Multi-media/multi-modal comparable corpora
 * Applications of Comparable Corpora
     - Human translations
     - Language learning
     - Cross-language information retrieval & document categorization
     - Bilingual projections
     - Machine translation
     - Writing assistance
 * Mining from Comparable Corpora
     - Extraction of parallel segments or paraphrases from
       comparable corpora
     - Extraction of bilingual and multilingual translations of
       single words and multi-word expressions; proper names, named
       entities, etc.


IMPORTANT DATES

 May 15, 2009 Paper submissions
 Jun 1, 2009 Notification of acceptance
 Jun 7, 2009 Camera-ready copies due
 Aug 6, 2009 Workshop date


SUBMISSION FORMAT

 Authors are invited to submit papers on original, unpublished work in
 the topic area of this workshop. We invite the presentation of:

    * Long papers should present completed work and should not exceed
      8 pages (plus one page of references);
    * Short papers can present work in progress (4 pages including
      references).

 Please use the official style files for ACL/IJCNLP 2009 available at:
 http://www.acl-ijcnlp-2009.org/main/authors/stylefiles/

 Submission site:
 https://www.softconf.com/acl-ijcnlp09/BUCC/

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


WORKSHOP CO-CHAIRS

 Pascale Fung, Hong Kong University of Science & Technology (HKUST)
 Pierre Zweigenbaum, LIMSI-CNRS (France)
 Reinhard Rapp, University of Mainz (Germany)
   and University of Tarragona (Spain)


PROGRAM COMMITTEE

 Hamdulla Askar(Xinjiang University, China)
 Srinivas Bangalore (AT&T Labs, US)
 Lynne Bowker (University of Ottawa, Canada)
 Éric Gaussier (Université Joseph Fourier, Grenoble, France)
 Gregory Grefenstette (Exalead, Paris, France)
 Hitoshi Isahara (National Institute of Information and Communications
 Technology, Japan)
 Min-Ye Kan (National University of Singapore)
 Adam Kilgarriff (Lexical Computing Ltd)
 Philippe Langlais (Université de Montréal, Canada)
 Rada Mihalcea (University of North Texas, US)
 Dragos Stefan Munteanu (Language Weaver, Inc., US)
 Grace Ngai (Hong Kong Polytechnic University, Hong Kong)
 Carol Peters (ISTI-CNR, Pisa, Italy)
 Serge Sharoff (University of Leeds, UK)
 Richard Sproat (OGI School of Science & Technology, US)
 Mandel Shi (Xiamen University, China)
 Yujie Zhang (National Institute of Information and Communications
 Technology, Japan)


WORKSHOP TECHNICAL SUPPORT

 Ricky Chan Ho Yin, Hong Kong University of Science & Technology