> Date: Thu, 7 Oct 2004 12:23:46 +0100
> From: Robert Sanderson <[log in to unmask]>
>
> Say I have the string:
> "XML in a Nutshell"
> and I have a system that knows XML --> Extensible Markup Language
>
> If a user searches for "Markup Language" then I should return a hit?
Yuck! This seems like just the find of feature that drives me crazy.
I would hate to use a server that did this kind of thing without being
explicitly asked to.
> Date: Fri, 8 Oct 2004 10:58:12 +0100
> From: Robert Sanderson <[log in to unmask]>
>
> > Does this generalise? i.e. would a modifier for equivalents work
> > here for both cases
> > Charles Lutwidge Dodgson being equivalent to Lewis Carroll
> > XML being equivalent to eXtensible Markup Language
>
> Looks good to me. cql.equivalent ?
I think that's _one_ of the things you'd want. But it's very much in
the same line to want to be able to search for "dinosaur" and find
records that contains "tyrannosaurus". And then of course you might
want to do the converse -- find "dinosaur" records when searching for
"tyrannosaurus". So that suggests we need at least three more
relation modifiers, and that in turn suggests that we shouldn't toss
them all into the lean, mean core "cql" set but instead make a new,
dedicated, set and put them in there:
fruitbat.equivalent
fruitbat.broader
fruitbat.narrower
and maybe others, such as fruitbat.related.
BTW., I am not claiming that "fruitbat" is necessarily the _very_ best
of all possible names for this putative new set.
_/|_ _______________________________________________________________
/o ) \/ Mike Taylor <[log in to unmask]> http://www.miketaylor.org.uk
)_v__/\ "It seems to be the word on the street that Hypsilophodontidae
is paraphyletic" -- Pete Buchholz.
--
Listen to free demos of soundtrack music for film, TV and radio
http://www.pipedreaming.org.uk/soundtrack/
