Corpora: ACL-02 Workshop CFP: Morphological and Phonological Learning

[Priscilla Rasmussen]

                     Workshop on
        Morphological and Phonological Learning

                  Philadelphia, PA
                    12 July 2001

                 Sixth Meeting of the
   ACL Special Interest Group in Computational Phonology
                in cooperation with
  ACL Special Interest Group in Natural Language Learning




                    Motivation

Two groups of researchers are converging on the need to construct
morphologies and phonologies of low density languages.

Natural language engineers hope to develop machine translation, speech
recognition, and other NLP technologies for these languages.  Meanwhile,
linguists and native speakers want to document the languages for
scientific or humanitarian reasons.  (This need is often expressed
concerning endangered languages, but is not restricted to that situation.)

This convergence of interests makes it an opportune time to meet to
discuss ways to analyze the morphology and phonology of a language (or a
group of related languages) more quickly (and perhaps more accurately)
than traditional methods have allowed.

Techniques for morphology and phonology learning may vary in the amount of
human involvement they require.  At one end of the spectrum are tools
intended to help a native speaker (perhaps with the aid of a linguist)
describe his or her own language.  At the other end are tools for
unsupervised machine learning from texts.  Intermediate or hybrid
approaches are also possible.

Methodologies to be discussed in this workshop need not be fully general:
for example, a tool might be best suited to agglutinating, fusional,
or polysynthetic languages, or specialized for compounding or
reduplication.

The Workshop on Morphological and Phonological Learning will be held July
12 2002, immediately after the ACL-02 meetings at the University of
Pennsylvania, Philadelphia Pennsylvania, USA.

More information about SIGPHON is available at
http://www.cogsci.ed.ac.uk/sigphon, and about SIGNLL at
http://ilk.kub.nl/~signll/.  The ACL-2002 website is
http://www.acl2002.org.




                Topics

o	Tools to help a native speaker or linguist visualize and describe
    the morphology and/or phonology of a language

o	Tools for (semi-)automated discovery of morphology and/or phonology

o	Databases and annotation tools designed for morphological or
    phonological information, particularly as these relate to learning

o	Resources for learning (taggers, seed grammars and lexicons, partially
    annotated text, bilingual text, etc.)

o	Linguistic (knowledge-based) approaches vs. empirical approaches;
    hybrid methodologies

o	Evaluation/comparison of morphology learning technologies

o	Adapting and reusing grammars and lexicons among related languages

o	Application of learned morphologies and phonologies (proofreading,
    machine translation, linguistic research, documentation of endangered
    languages, speech recognition)

o   Theoretical results on learnability or representation






            Program Committee

o   Mike Maxwell (Linguistic Data Consortium, chair)
o   Antal van den Bosch (Tilburg University, SIGNLL)
o   Jason Eisner (Johns Hopkins University)
o   Steven Bird (University of Pennsylvania)
o   Lauri Karttunen (Parc Inc.)
o   John A. Goldsmith (University of Chicago)

Invited Speaker: David Yarowsky, Johns Hopkins University




                Submission Format and Instructions

Submissions must be in English, and should be full-length papers, up to a
maximum of 10 pages. (The final version in the proceedings should
incorporate reviewers' suggestions and may be up to 12 pages.) Except for
length, papers should adhere to the two-column format prescribed by
ACL'2002 Please see http://www.cs.ualberta.ca/~lindek/acl02/style/ for the
detailed guidelines.

Submissions should be sent electronically in Word, PDF or ASCII text
format to arrive no later than April 15, 2002 to Mike Maxwell
(maxwell at ldc.upenn.edu).  The body of the email should give title,
author(s), abstract, and contact information.  The subject line
should be "Morphology/ Phonology Learning Workshop."

The workshop papers will be published in both electronic and hard-copy
form.

While electronic submission is strongly preferred, if for some reason you
are not able to submit electronically, contact Mike Maxwell concerning
hard-copy submission.

Completed work is preferable to intended work, but in any event the paper
should clearly indicate the state of completion of the reported results.
Submissions will be evaluated on the basis of their relevance, innovation,
quality, and presentation according to the schedule below.  The program
chair may invite additional reviewers as necessary to obtain relevant
expertise and avoid conflicts of interest.

Authors will be expected to archive their final papers in the Computation
and Language section of arXiv.org, and to submit pointers to the archived
paper when submitting their camera-ready copy.



                Schedule

o   Submission Deadline:       5 April 2002
o   Notification:             25 April 2002
o   Camera-ready Copy Due:    21 May 2002
o   Workshop:                 12 July 2002