Re: Indexing E. documents

Subject: Re: Indexing E. documents
From: "Susan Self @ignite" <susan -at- ALSYS -dot- COM>
Date: Fri, 24 May 1996 12:07:05 PDT

David Blyth (Hi, David!) wrote:

>1) Has anyone read the STC article on indexing Electronic documents?

>2) Any reactions?

>3) Does anyone have any other idea on indexing Electronic documents?
> On the Web?

>I've read the article and have a few other ideas, but find it a
>difficult topic. So I thought we could discuss the subject then
>I would steal thoughts.... ;)

I read Lori Lathrop's article (STC Intercom, March 1996) and found
the information very similar to what she presents in her indexing
seminars. I feel the physical limitations of the online medium
demand more detailed attention than the article provided.

For instance, the problem that ravages me the most is the access
to the index. Frame supports hypertext links for each letter of the
alphabet in a row at the top of a page. I suppose that one could
implement the same for online indexes. The problem is that any good
index needs to be rather long to be thorough and useful. For instance,
I am going through my indexes and making separate subentries for
every entry that has more than one page reference. The result is
that every entry and subentry has a descriptive phrase distinguishing
it from other similar entries and only one page reference. When
such an index goes online, the page references can become links.
The user doesn't have to try three different links for one entry
to see which page has the desired information. The descriptive
phrase should do the discriminating beforehand. With this method,
one four hundred-page manual has a thirty-three-page index.
With five manuals online and a global index that covers all of them,
you can imagine the global index being booklength in itself.
How to navigate such an index, and whether it will be so cumbersome
for users as to be useless, are the questions that I can't answer yet.
I need to try the implementation and see how much quantity
destroys quality and usefulness.

That's why I'm investigating online search engines based on the UNIX
grep command, since my target audience are UNIX users. Users can type
a word in an HTML form and produce another HTML file with links to
every use of the word within the document search space in the context
of a sentence. Clicking on a link takes them to the line in the file
referenced. Users might find this method more useful in cases where
they have a specific term in mind that is not used too frequently.
Of course, it can never take the place of the more conceptual groupings
and hierarchies of a good index, but it may provide faster navigation
for certain types of lookups.

Susan Self
susan -at- thomsoft -dot- com

^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Post Message: TECHWR-L -at- LISTSERV -dot- OKSTATE -dot- EDU
Get Commands: LISTSERV -at- LISTSERV -dot- OKSTATE -dot- EDU with "help" in body.
Unsubscribe: LISTSERV -at- LISTSERV -dot- OKSTATE -dot- EDU with "signoff TECHWR-L"
Listowner: ejray -at- ionet -dot- net


Previous by Author: Re: Functionality
Next by Author: Re: Indexing E. documents
Previous by Thread: Re: Indexing E. documents
Next by Thread: Re: Indexing E. documents


What this post helpful? Share it with friends and colleagues:


Sponsored Ads