Re: Creating Indexes

[pete swisher <PSwisher -at- QUARK -dot- COM>]

Quark does indeed have index XTensions software.  This tool makes it very
easy to tag words for an index, specify their "level", and determine their
relation to other words in an index.  For more info:
http://www.quark.com/files/xtquarkxts_40.html

Pete
> -----Original Message-----
> From: David Downing [SMTP:DavidDowning -at- USERS -dot- COM]
> Sent: Monday, March 22, 1999 10:53 AM
> To:   TECHWR-L -at- LISTSERV -dot- OKSTATE -dot- EDU
> Subject:      Re: Creating Indexes
>
> Your question seemed a bit open-ended, so I hope I'm addressing the point
> you wanted addressed.
>
> It sounds like you're looking for a tool that'll completely automate the
> task of indexing your documents.  Well, I'm a former technical services
> librarian, and at the risk of sounding like an old fuddy-duddy, I don't
> think totally automated indexing -- going through a document for every
> instance of specified words and phrases -- is ever going to be totally
> successful..  The reason is that the use of a given word or phrase doesn't
> always equate with a reference to a given subject.
>
> For example, consider the following two statements:
>
> Error messages must be no more than ten words long.
>
> If you see a series of error messages, you should stop the installation
> process immediately.
>
> Both statements use the phrase "error messages," but only the first one
> actually * discusses * error messages.  Furthermore, a person seeking the
> information in the first statement is likely to look under "Length of
> error
> messages," "Error messages, length of," or some variation that involves
> the
> word "length" -- which the statement doesn't actually use.
>
> I'm more comfortable with what I call the "semi-automated" method, in
> which
> the writer embeds terms in the document and the indexing software searches
> for the embedded terms.  Since writers can do this as they're writing,
> it's
> not a separate step, as it would be if they were doing it manually, and
> when
> documents are revised, the process is almost totally automated because the
> terms are already embedded.
>
> Admittedly, the non-correspondence between words/phrases and topic
> references is less of a problem in technical documents than in other types
> of writing, but it's still there.  It's inherent in the use of natural
> language.
>
>         David
>
> > Subject:      Creating Indexes
> >
> > We would like to improve the quality of these indexes and would
> > like to find out any strategies or special software technical writers
> and
> > editors may use.
> >
> > editors comb through the Quarks by hand
> > Any leads on a better method would be greatly appreciated.
> >
> >
>
> From ??? -at- ??? Sun Jan 00 00:00:00 0000=
> =
>
>

 From ??? -at- ??? Sun Jan 00 00:00:00 0000=