Following the success of the first Workshop on Building and Using 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.
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.
Kenneth Ward Church (Microsoft Research, Redmond)
We solicit contributions in but not limited to the following topics:
| Call for papers | Feb 6, 2009 |
| Paper submissions | May 1, 2009 |
| Notification of acceptance | Jun 1, 2009 |
| Camera-ready copies due | Jun 7, 2009 |
| Workshop date | Aug 6, 2009 |
Authors are invited to submit papers on original, unpublished work in the topic area of this workshop. We invite the presentation of:
All submissions are to be formatted using the ACL/IJCNLP 2009 style files available at: http://www.acl-ijcnlp-2009.org/main/authors/stylefiles/
To submit a paper, go to the submission page at: https://www.softconf.com/acl-ijcnlp09/BUCC/
The deadline for paper submission is at 24:00 GMT on May 15, 2009.
Authors may submit the same paper at several conferences. In this case they must notify the organizers in a separate mail to pascale at ee dot ust dot hk, so we know that the paper might be withdrawn depending on the results at some other conference.
Withdrawals should be notified by email to the organizers by June 2nd.
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)
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)
Satoshi 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)
Carole 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)