Background and Goals

Growing interest in Chinese language processing is leading to the
development of resources such as annotated corpora, word segmenters,
part-of-speech taggers, and parsers. As more resources have become
available recently, it is crucial to create a platform that allows
easy exchange of information and data and the comparison of different
approaches to various NLP tasks. The SIGHAN workshops provide a forum
where the latest research in these areas can be shared.

Past SIGHAN workshops included the organization of the First and
Second International Chinese Word Segmentation Bakeoff, where many
word segmentation systems from academia and industry were
evaluated. The evaluations conducted have proven to be influential,
and the evaluation data set has become the benchmark for Chinese word
segmentation in the Chinese language processing community.

COLING/ACL 2006 in Sydney will provide an ideal opportunity to bring
together again influential as well as aspiring researchers from Hong
Kong, Mainland China, Singapore, and Taiwan and other interested
Chinese language processing researchers from around the world, to
deliberate and interact on a range of NLP issues.

The first day of the workshop (July 22) will consist of papers on all
aspects of Chinese language processing, including but not limited to:

• word segmentation
• part-of-speech tagging
• parsing
• lexical semantics
• word sense disambiguation
• lexicon acquisition
• corpus development
• discourse processing
• generation
• cross-lingual information retrieval
• machine translation

The second half-day of the workshop (July 23) will present results
from a bakeoff. A SIGHAN business meeting will discuss lessons learned
and future plans. This year's third SIGHAN bakeoff will be held during
the Spring of 2006. This year, in addition to the standard Chinese
word segmentation task, we also plan a new track on named entity
recognition and tagging from unsegmented Chinese text. Training and
testing resources will be provided in both traditional and simplified
character sets from a range of institutions including Chinese
Knowledge and Information Processing group of Academia Sinica, City
University of Hong Kong, Microsoft Research Asia, Peking University,
and the Universities of Pennsylvania and Colorado.

Please also check the official Bakeoff website.

Submission Method

Papers should be written in English and may not exceed 8 pages
(including all illustrations, references and appendices, and using
11pt for the main text). We strongly recommend the use of the LaTeX
style files or MS Word document template provided by COLING/ACL 2006,
available at http://www.acl2006.org/program/style. Since reviewing
will be blind, manuscripts should not include authors' names and
affiliations. Papers should be submitted via our electronic submission site.

Important Dates

Workshop paper submission deadline:		April 12, 2006
Notification of acceptance:		May 12, 2006
Camera ready version deadline:		May 31, 2006