I assume I would use Perl to parse the newsgroups and use lisp to
apply the AI that filters the spam. I would love to hear from someone
who knows lisp, perl and prolog. T.I.A.
-RB
Fight Censorship
From: Reini Urban
Subject: Re: Compare Lisp To Perl - Sample Project: Anti-Spam Newsreader
Date:
Message-ID: <35c26a4b.8488505@judy>
······@aol.com (Recordable Bob) wrote:
>I assume I would use Perl to parse the newsgroups and use lisp to
>apply the AI that filters the spam. I would love to hear from someone
>who knows lisp, perl and prolog. T.I.A.
I would either do it in lisp or perl alone, but not in prolog. maybe
awk. I would say perl, because of the simple regex with search/replace
and you'll get some antispam regex'ed lists on the net, you could use
immediatly in perl or awk, but would need some conversion for lisp.
there are regex libs for lisp but in perl it's builtin. (based in the
same spencer code mostly)
there's not that much AI in anti-spam filtering code that lisp should be
preferred. but if then use lisp alone. maintaining modules in seperate
languages makes your head aching, esp. with perl or c :)
--
Reini Urban
From: David Bakhash
Subject: Re: Compare Lisp To Perl - Sample Project: Anti-Spam Newsreader
Date:
Message-ID: <cxj7m0sit88.fsf@engc.bu.edu>
······@aol.com (Recordable Bob) writes:
> I assume I would use Perl to parse the newsgroups and use lisp to
> apply the AI that filters the spam. I would love to hear from someone
> who knows lisp, perl and prolog. T.I.A.
I wrote a Lisp program to do this from start to finish, and did it for
someone who was writing a book. I charged him $1100 for use of the
program. It went from gathering the posts (from a very large,
long-term database) and after filtering, sorting, threading, etc,
compiled the new database. If you want, I can sell you the program.
In order to use it, all you need to know is what you're searching
for. the program does the rest. I've upgraded the program since, and
so I can sell you the use of it for a comprable price (maybe a bit
more, if you don't mind, since it's a bit better now). You can email
me back and let me know.
I don't use prolog, but rather rather other contstraints packages
(check out Oz). But that is not necessary for such a problem.
lastly, I am a Lisp and Perl programmer, and Lisp is far superior for
this problem; otherwise, I'd have done it in Perl. If anyone says
otherwise, then I'll contest it. However, I can see how a Perl
version would be faster than mine. I don't think speed is critical
though, since the bulk of the work is the network latency, and
downloading.
dave