If you want more information about the software, please send a
message to one of the following addresses:
ARPA: con.r...@su-gsb-how.ARPA
UUCP: ucbvax!con.r...@su-gsb-how.ARPA
Thanks much,
Brian Reilly
--
--rick heli
UUCP: ... {ucbvax,lll-lcc}!ucdavis!ccrrick
INTERNET: ucdavis!ccr...@ucbvax.BERKELEY.EDU
I can think of lots of other uses for such computer-readable text. We just
spent a man-day or so typing in the front page of the New York Times, so we
could run our parser program on it (as a test case). (We would probably have
done better if we got a secretary to do it, but we didn't realize it would be
such a big task.) I know of lots of sources of such data that you can spend
big bucks on, but there's probably some public domain stuff out there...
BTW, we decided against usnet as a source of text because we wanted something
at a slightly higher level :-)
All seriousness aside, you might consider some of the discussion on
net.origins; some of the articles there are certainly long enough...
(And yes, I do mean "all seriousness aside".)
--
Mike Maxwell
Boeing Artificial Intelligence Center
...uw-beaver!uw-june!bcsaic!michaelm