<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
</head>
<body bgcolor="#ffffff" text="#000000">
<br>
<div class="moz-text-flowed"
style="font-family: -moz-fixed; font-size: 13px;" lang="x-western">
CALL FOR PAPERS<br>
<br>
Fourth Workshop on<br>
Syntax and Structure in Statistical Translation<br>
(SSST-4)<br>
<br>
COLING 2010 / SIGMT Workshop<br>
28 August 2010, Beijing<br>
<br>
*** Submission deadline: 25 Jun 2010 ***<br>
*** Invited Keynote: Martin Kay ***<br>
<br>
The Fourth Workshop on Syntax and Structure in Statistical Translation<br>
(SSST-4) seeks to build on the foundations established in the first
three<br>
SSST workshops, which brought together a large number of researchers<br>
working on diverse aspects of synchronous/transduction grammars<br>
(hereafter, S/TGs) in relation to statistical machine translation. Its<br>
program each year has comprised high-quality papers discussing current<br>
work spanning topics including: new grammatical models of translation;<br>
new learning methods for syntax-based models; using S/TGs for semantics<br>
and generation; syntax-based evaluation of machine translation; and<br>
formal properties of S/TGs. Presentations have led to productive and<br>
thought-provoking discussions, comparing and contrasting different<br>
approaches, and identifying the questions that are most pressing for<br>
future progress in this topic.<br>
<br>
The need for structural mappings between languages is widely recognized<br>
in the fields of statistical machine translation and spoken language<br>
translation, and there is a growing consensus that these mappings are<br>
appropriately represented using a family of formalisms that includes<br>
synchronous/transduction grammars and their tree-transducer equivalents.<br>
To date, flat-structured models, such as the word-based IBM models of<br>
the early 1990s or the more recent phrase-based models, remain widely<br>
used. But tree-structured mappings arguably offer a much greater<br>
potential for learning valid generalizations about relationships between<br>
languages.<br>
<br>
Within this area of research there is a rich diversity of approaches.<br>
There is active research ranging from formal properties of S/TGs to<br>
large-scale end-to-end systems. There are approaches that make heavy use<br>
of linguistic theory, and approaches that use little or none. There is<br>
theoretical work characterizing the expressiveness and complexity of<br>
particular formalisms, as well as empirical work assessing their<br>
modeling accuracy and descriptive adequacy across various language<br>
pairs. There is work being done to invent better translation models, and<br>
work to design better algorithms. Recent years have seen significant<br>
progress on all these fronts. In particular, systems based on these<br>
formalisms are now top contenders in MT evaluations.<br>
<br>
We invite papers on:<br>
<br>
* syntax-based / tree-structured statistical translation models and<br>
language models<br>
* machine learning techniques for inducing structured translation
models<br>
* algorithms for training, decoding, and scoring with S/TGs<br>
* empirical studies on adequacy and efficiency of formalisms<br>
* studies on the usefulness of syntactic resources for translation<br>
* formal properties of S/TGs<br>
* scalability of structured translation methods to small or large
data<br>
* applications of S/TGs to related areas including:<br>
o speech translation<br>
o formal semantics and semantic parsing<br>
o paraphrases and textual entailment<br>
o information retrieval and extraction<br>
<br>
For more information: <a class="moz-txt-link-freetext" href="http://www.cs.ust.hk/~dekai/ssst/">http://www.cs.ust.hk/~dekai/ssst/</a><br>
<br>
<br>
INVITED KEYNOTE<br>
<br>
Martin Kay<br>
<br>
<br>
IMPORTANT DATES<br>
<br>
Submission deadline: 25 Jun 2010<br>
Notification to authors: 9 Jul 2010<br>
Camera copy deadline: 18 Jul 2010<br>
<br>
<br>
SUBMISSION<br>
<br>
Papers will be accepted on or before 25 Jun 2010 in PDF or Postscript<br>
formats via the START system at<br>
<a class="moz-txt-link-freetext" href="https://www.softconf.com/coling2010/SSST2010/">https://www.softconf.com/coling2010/SSST2010/</a>. Submissions should follow<br>
the COLING 2010 length and formatting requirements for full papers of<br>
eight (8) pages of content with one (1) extra page for references, found<br>
at <a class="moz-txt-link-freetext" href="http://www.coling-2010.org/SubmissionGuideline.htm">http://www.coling-2010.org/SubmissionGuideline.htm</a>.<br>
<br>
<br>
CAMERA COPY<br>
<br>
Camera ready final versions will be accepted on or before 18 Jul 2010 in<br>
PDF or Postscript formats via the START system at<br>
<a class="moz-txt-link-freetext" href="https://www.softconf.com/coling2010/SSST2010/">https://www.softconf.com/coling2010/SSST2010/</a>. Papers should still
follow<br>
the NAACL HLT 2009 length and formatting requirements for full papers,<br>
found at <a class="moz-txt-link-freetext" href="http://www.coling-2010.org/SubmissionGuideline.htm">http://www.coling-2010.org/SubmissionGuideline.htm</a>.<br>
<br>
<br>
CONTACT<br>
<br>
Please send inquiries to <a class="moz-txt-link-abbreviated" href="mailto:ssst@cs.ust.hk">ssst@cs.ust.hk</a>.<br>
<br>
</div>
</body>
</html>