[
Lists Home |
Date Index |
Thread Index
]
[Micah Dubinko]
> >Two databases may
> >use different identifiers for what is in fact the same concept, such as
> >zip code. A program that wants to compare or combine information across
> >the two databases has to know that these two terms are being used to
> >mean the same thing.
>
> Isn't that the whole crux of the problem? If the problem is that two
> databases (or web pages) use a different name for "zip code", the solution
> is to add a third (Q)name, that both must agree upon?
>
Not necessarily. Another approach is to declare a mapping between the two
data items with different names. And one way to do that is to make an
assertion that they are equivalent - like {database1:zip_code,
isEquivalentTo, database2:postal_code}. Naturally, the semantics for
isEquivalentTo has to be established in some way. RDF/DAML_OIL/WEBONT
provide a mechanism to exchange such assertions, and inference engines can
make use of them (and the same for Topic Maps, which are also being used for
inter-database integration via mapping). Of course, simple mappings do not
need inference engines, they can be used more directly.
This approach can be used where there is no way for the parties to agree on
a common name and change their databases to use it.
Cheers,
Tom P
|
