|
|
Sigfrid Lundberg, Lub NetLab <[log in to unmask]> wrote:
>> what about a more general algorithm for dumb down
>> of structured values:
> I like Stefan's algorithm. I agree with others here who have remarked
> that this isn't really the same thing as the string you would extract
> from a structure by removing all encoding schemes and refinements.
But remember -- encoding schemes and refinements are for moving through
"is-a" links. Stefan's proposal is for moving through "has-a" links.
> We need to be able to search for both when searching for an
> author. (When it comes to querying a triple database I'm still a
> newbie).
This is an easy software issue to resolve. (Although it involves a tradeoff
between speed and storage space...)
--
[ Aaron Swartz | [log in to unmask] | http://www.aaronsw.com ]
|
|
|
|