Re: virus: Archives
red_mist (red_mist@portsurfer.demon.co.uk)
Mon, 11 May 1998 23:24:10 +0100
In message <3.0.3.32.19980510163857.007d74b0@mail.rdc1.bc.wave.home.com>
, Dan Plante <dplante@home.com> writes
> Well, that's not what I had in mind. Maybe the reference to the unix
>"grep" command (and the Windows version, WinGrep) was a little obscure.
>Grep is a script tool that gives great flexibility in searching and
>formatting output from multiple files based on a myriad of operators
>called "regular expressions" and was usually sourced from GNU works:
>Gnu Regular ExPressions, or GREP.
>
> I was more interested in crunching the archives to extract and condense
>the archives based on topic, to see how (or if) the prevailing opinions
>on any topic changed over time. This is difficult unless you have a tool
>that allows you to deduce the real topic of conversation by checking the
>context of the keywords using nearness operators, among others.
>
Yeah i used it before but it would be a little difficult to format that
into a cgi script to put on the server. I haven't really experimented
much with it so I'll see what I can do. Also there would have to be
some type of thing that rates the articals as to how they best fit the
query, like a lot of the search engines on the web. I'll do a bit of
experimenting withi it and see what can be done. Also a rating scheme
would be good to have as well
--
Only the weak are blind when the mist descends
red_mist