LINGUIST List 9.188

Sat Feb 7 1998

FYI: NSF, WWW Semantic Net, Phonology Game, ETRW

Editor for this issue: Martin Jacobsen <martylinguistlist.org>

Directory

1. pchapin, Major new NSF funding initiative
2. Douglas Beeferman, WWW Semantic Net Resource
3. Sean Jensen, The Phonology Game
4. Helmer Strik, ETRW 'Modeling pronunciation variation for ASR'

The U.S. National Science Foundation has announced a new
multidisciplinary funding initiative in Knowledge and Distributed
Intelligence (KDI). KDI includes a second competition for grants for
research in Learning and Intelligent Systems (LIS - an activity which
began in 1997), as well as other components. NSF will award
approximately $50 million in 1998 under this initiative. Letters of
intent to apply are due at NSF by April 1, and full proposals are due
by May 8. Only researchers affiliated with U.S. institutions are
eligible to apply.

The full proposal solicitation for KDI is posted on the NSF Web site
at

 http://www.nsf.gov/kdi

A brief overview of the initiative appears below.

Paul Chapin, U.S. National Science Foundation
*********************************************
Knowledge and Distributed Intelligence (KDI)

The recent explosive growth in computer power and connectivity is
reshaping relationships among people and organizations, and
transforming the processes of discovery, learning, and communication.
As a result of the technological advances we have unprecedented
opportunities for providing rapid and efficient access to enormous
amounts of knowledge and information; for studying vastly more complex
systems than was hitherto possible; and for advancing in fundamental
ways our understanding of learning and intelligent behavior in living
and engineered systems. NSF's Knowledge and Distributed Intelligence
(KDI) theme is a Foundation-wide effort to promote the realization of
these opportunities.

Three Foci for FY 1998: KN, LIS, and NCC

To achieve the aims of KDI, proposals are solicited from individuals
or groups for research that is inherently multidisciplinary or that,
while lying within a single discipline, has clear impact on at least
one other discipline. In FY 1998, KDI will have three foci: Knowledge
Networking (KN); Learning and Intelligent Systems (LIS); and New
Computational Challenges (NCC).

Knowledge Networking (KN) focuses on the integration of knowledge from
different sources and domains across space and time. The goal of KN
research is to achieve new levels of knowledge integration,
information flow, and interactivity among people, organizations, and
communities, and to deepen our understanding of the ethical, legal,
and social implications of knowledge networking.

Learning and Intelligent Systems (LIS), an ongoing program, seeks to
stimulate multidisciplinary research that will unify experimentally
and theoretically derived concepts related to learning and intelligent
systems, and that will promote the use and development of information
technologies in learning and discovery across a wide variety of
fields. LIS emphasizes research that advances basic understanding of
learning and intelligence in natural and artificial systems, as well
as research that supports the development of tools and environments to
test and apply this understanding in real situations.

New Computational Challenges (NCC) focuses on research and tools
needed to discover, model, simulate, analyze, display, or understand
complicated phenomena, to control resources and deal with massive
volumes of data in real time, and to predict the behavior of complex
systems. These aims will require major advances in hardware and
software to handle complexity, representation, and scale, to enable
distributed collaboration, and to facilitate real-time interactions
and control.

The KDI Competition

A KDI proposal solicitation has just been released; the full text is
posted on the KDI web page at http://www.nsf.gov/kdi. Letters of
intent are due April 1, 1998, and the deadline for full proposals is
May 8. Awards will be made in the fall. Approximately $50 million is
available for funding proposals submitted to this competition.

Proposals are solicited for any amount up to $1.0 million per year for
up to three years. We expect to make grants at a wide variety of
amounts and durations. In exceptional cases, awards for up to five
years may be considered if the justification and promise are
compelling.

A second KDI competition will be held in FY 1999, subject to
availability of funds. An updated solicitation, which may include
revised research emphases, will be released in advance of this
competition.

For more details please see the solicitation.

Now available on the Web for your entertainment and edification is a
semantic network program called Lexical FreeNet:

	http://www.link.cs.cmu.edu/lexfn/

FreeNet is a database system for creating and querying "finite
relation expression networks". Lexical FreeNet combines several
WordNet-derived relations with a statistically-derived trigger
relation, and throws in a couple phonetic dictionary-derived relations
for good measure. The result is, I hope, a useful resource for
discovering relationships between concepts across dimensions of
meaning and sound. You can presently issue shortest path queries,
concept "intersection" queries, and "rhyme coercion" queries. The
latter finds rhyming word pairs that are similar in meaning to two
input concepts. You can also use the system simply to lookup the
words that are related in some way to a source, or find words that are
spelled similarly to a character sequence.

My intent is to use the system for continued research on automatic
segmentation, summarization, and information extraction on transcripts
of spontaneous speech. The source code is not yet publically
available but if there is sufficient interest I would be happy to make
it so.

I think this program has the potential to be valuable in an
educational setting. Please feel free to check out the program and
read the technical note and usage notes if you get a chance. Any
feedback is much appreciated.


Regards,
Doug

Dear fellow LINGUISTs

Here is a party-game, primarily for phonologists, but morphologists
can play too. Each player picks any number of phonological (or
morphological) rules they like. Then each player lists all the
exceptions (lexical exceptions, readjustments, loanwords etc, etc.) to
those rules. The winner is the player holding a rule with no
exceptions. I have been playing the game for years, and have still to
win.

Can I ask LINGUIST subscribers in an idle moment to play a round or
two and email me their WINNING hands. I will post a summary, but my
guess is it will be VERY short. Winners, if there are any, will be
able to go on to Round 2 of the game, which I will announce if there
is sufficient interest or playfulness out there. I may even be
persuaded to offer a modest prize to the ultimate winners.

Players in the OT camp can by definition never win, of course, because
ALL their rules are violable :(

Best of luck!

Sean.

- --
Sean Jensen
e-mail: seanjseanj.demon.co.uk
www: http://www.seanj.demon.co.uk
tel/fax: +86 20 8759 6961
snail: Unit 13b, Block 2, The Greenery, 55-79 Huayang Jie, Tiyu Dong
Lu, Tianhe, Guangzhou 510620, China

 Call for Participation

In this e-mail you will find some (links to) new information regarding
the ESCA Tutorial and Research Workshop 'Modeling pronunciation
variation for automatic speech recognition'.

Most of the information can be found at the following sites:

* http://lands.let.kun.nl/pron-var/
workshop home page

* http://lands.let.kun.nl/TSpublic/strik/pron-var/ws-papers.html
abstracts of accepted papers

* http://lands.let.kun.nl/TSpublic/strik/info-ned.html
information for planning your trip to the workshop

* http://lands.let.kun.nl/pron-var/registration.html
details about registration


The deadline for early registration is 16-02-98.

It is possible to attend the workshop without presenting a paper.
However, attendence will be limited for two reasons. First of all,
because we want to stimulate discussions. And second, because of
conference facility constraints.

Furthermore, priority will be given to authors presenting papers at
the workshop and members of the COST Telecom action 249 (COST 249 is
one of the organizers of this workshop, and immediately following this
workshop there will be a COST 249 meeting). Finally, the registration
fee is increased with 200 NLG after 16-02-1998. That is why potential
attendees are highly encouraged to submit a registration form as soon
as possible.

For registration see
http://lands.let.kun.nl/pron-var/registration.html 
or contact:

 WORKSHOP SECRETARIAT

 Marie-Jose Verbeek
 Bureau van de pedel
 Comeniuslaan 2
 P.O. Box 9111
 6500 HN Nijmegen
 the Netherlands

 Tel.nr.: 31-24-3615968
 Fax nr.: 31-24-3567956 / 3616135
 E-mail : M.Verbeekburo.kun.nl


With kind regards,
Helmer Strik

Dr. H. Strik Dept. of Language and Speech
 P.O. Box 9103
Tel.nr.: +31-24-3616104 6500 HD Nijmegen
Fax nr.: +31-24-3612907 The Netherlands
E-mail : STRIKLET.KUN.NL
URL http://lands.let.kun.nl/TSpublic/strik/