home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!ulowell!m2c!bu.edu!purdue!haven.umd.edu!mimsy!terry
- From: terry@cs.umd.edu (Terry Gaasterland)
- Newsgroups: comp.ai
- Subject: Re: Natural Language newsgroup
- Message-ID: <63725@mimsy.umd.edu>
- Date: 26 Jan 93 18:33:18 GMT
- References: <1jvq1jINNidk@mojo.eng.umd.edu>
- Sender: news@mimsy.umd.edu
- Organization: Math and CS Division, Argonne National Lab, Illinois
- Lines: 81
-
- In article <1jvq1jINNidk@mojo.eng.umd.edu> clin@eng.umd.edu (Charles Lin) writes:
- >***
- >Charter for Natural Language Newsgroup
-
- It is important to differentiate between a comp.ai.natural-lang
- group and existing groups: e.g. sci.lang, the nl-kr group (I
- forget its newsgroup name), and the LINGUIST list. What would
- a comp.ai.natural-lang group provide that these groups do not?
-
- NL-KR is carefully moderated. As a result, it is not so much a forum
- for dialog as a place for announcements and requests for information.
- The main problem with the NL-KR group is that people ask questions
- and others answer them privately. Many times over the years, I have
- sent people requests for a copy of the responses that they get ---
- usually to no avail.
-
- The sci.lang group has a general scope that basically covers all
- issues in linguistics. It tends to get a lot of noise. I would hope
- that a comp.ai.natural-lang group would successfully focus on the
- issues involved in getting computers to deal with processing language.
-
- The LINGUIST list is a forum for linguists to exchange views, argue
- about points, give circumstantial data points to each other, and
- post announcements. The focus is on linguistics, not on how to
- get computers to deal with language. Its very high volume of
- messages reflects the size of the readership and its willingness
- to participate in discussions. The list is a good place for a
- computer scientist to pick up a general sense of issues and
- topics in linguistics. Because it is lightly moderated, the
- LINGUIST list is less noisy than the sci.lang group. Because
- of the quality of the list members, the discussion tends to
- stay at a professional, scientific level. These properties
- are harder to acheive with an unmoderated newsgroup.
-
- The issues that are dealt with in (computer oriented) natural
- language research lie at the heart of AI. A comp.ai.natural-lang
- group should provide a forem for specific, detailed "intra-disciplinary"
- discussion (intra-AI, that is) that explores for example, the
- relationships between language learning and visual recognition,
- language understanding and visual perception, memory and language
- generation, knowledge representation and language generation,
- planning and language generation, spatial representation and
- locative prepositions, temporal representation and tense and
- aspect, to name just a few current areas of active investigation.
-
- Comments? Suggestions?
-
- > Name: comp.ai.natural-lang
- > Moderated?: Unmoderated
- > Purpose: To discuss issues related to natural language, especially
- > but not solely the kind of issues that researchers in
- > AI would consider important to natural language. This
- > would include issues such as:
- >
- > * Natural Language Understanding
- > * Natural Language Generation
- > * Machine Translation
- >
- > Issues related to specific research being done in the
- > field may also be considered.
- >
- >***
- >
- > If anyone else has any additions, please contact me at clin@eng.umd.edu.
- >Jaime Cisneros has expressed interest in helping out, but to make it
- >easier, I'll use my e-mail address. If this bounces, you can try
- >him at cisneros@eola.cs.ucf.edu...
- > ...
- >--
- >Charles Lin
- >clin@eng.umd.edu
-
-
- I think that the clarifications above, or some instantiation of them,
- should be made part of the charter so that unfocused, unprofessional,
- and/or rampantly unscientific discussion can be moderated through
- peer pressure with references to the charter.
-
- -- Terry Gaasterland
-
- (gaasterland@mcs.anl.gov or terry@cs.umd.edu)
-