[Sigia-l] search results and thesauri

[Andrew McNaughton]

On Sat, 25 May 2002, Ziya Oz wrote:

> Fortunately there's a way to perform arbitrary set operations on clusters of
> taxonomic terms or keywords *without* having the database to do any search
> until the user is satisfied with his criteria.
>
> In other words, if the user sees that a search on 'word_1' would produce a
> million records, it's probably not going to be very useful. So why do the
> search to begin with, wasting the user's time, taxing the server/DB and
> clogging the network?
>
> At that point, the application should be set to offer 'related' words, each
> indicating how many results it would produce. The user can then AND/OR these
> words and immediately see the size of the resulting set, hopefully
> approaching a reasonable, small number.

Any examples?  I'd be curious to see what the interface for this might be
like.  Designing Web interfaces for defining boolean query structure is
kind of clumsy, and I'd like to see any good examples you or others know
about.

One imaginative solution know about is aidministrator.nl's Cluster maps.
I'd give you an URL except that it's so god-awful you don't want to know.
www.aidministrator.nl -> Spectacle -> Cluster Maps.

It appears that  these are intended to be used within specialised client
software, but the basic visualisation idea could easily be adapted
(possible intellectual property issues aside).

> Sets are pre-calculated and thus set operations are instantaneous, requiring
> zero search of actual records. Once a small number is reached, a button
> would perform the actual search across records.

This works for a very small number of terms, but the server load and page
size required to deliver this pre-calculated info to a client side tool
increases dramatically with the number of terms involved.  If all
combinations are pre-calculated then it sounds workable for three or four
terms, dubious for 5 or 6, and most unlikely for 7 or 8 terms.

Andrew McNaughton