RE: Consolidating RDF XML data with different ids

Hi David,
Please find my comments below, marked [HT]
Regards,
Hans

____________________
Hans Teijgeler
ISO 15926 specialist
Netherlands
+31-72-509 2005
www.InfowebML.ws <http://www.infowebml.ws/>  
hans.teijgeler@quicknet.nl

-----Original Message-----
From: semantic-web-request@w3.org [ <mailto:semantic-web-request@w3.org>
mailto:semantic-web-request@w3.org] On Behalf Of David Pratt
Sent: Sunday, January 22, 2006 5:57
To: semantic-web@w3.org
Subject: Consolidating RDF XML data with different ids



Hi. I am trying to work through some practical aspects of RDF XML. I am
hoping to receive advice to a hypothetical situation so that I have a
clearer idea of how one might go about consolidating data from more than
one graph into a single graph in a reasonable way.

Let's say I have a schema for automobiles where there are a number of
properties that describe each car including model, manufacuter etc. I
also have a schema for car parts also.

The schema is used by more than one person at different locations that
have no communication with each other. Each automobile has a id.

[HT] Be precise: here you mean that each automibile *class* has an id

In addition to this when an automobile record was created, if a automobile
manufacturer was not found in the graph, new triples were also created
for the manufacturer which is also given a unique id.

[HT] That manufacturer would be an individual

Now I want to consolidate my automobile graphs from different locations
  into a national data store where I can query for any car from my
vendors. I find that several people have identified the same
manufacturer but using different ids.

[HT] No problem: use owl:sameAs [1]

I also find that essentially the
same car has been described at more than one location for the exception
of its vehicle identification number (its id) with more or less detail
from different dealers. 

[HT] That 'vehicle identification number' is a Property of a *member* of
your car class.

To further complicate matters I sell car parts,
and my parts dealers also have assigned different id's to the same
generic part, let's say a muffler for a 1990 Chev truck.

{HT] That doesn't complicate it, it is more of the same problem

It turns out that some vendors are more descriptive (using more
properties in the schema) than others. When I consolidate, I would
really like to combine the efforts of the vendors so that my part
information in my national data store is as comprehensive as possible.

[HT] You can only safely consolidate that IFF the more and the less
described classes are *really* the same. In that case use
owl:equivalentClass [2]

Given this scenario, what would be the best approach for consolidating
the information as much as possible. I would appreciate comments on how
one might accomplish this in a way that may not produce much unnecessary
duplication in the data store. How would I deal with different ids given
to say Ford Motor Company as a manufacturer by different vendors yet
they all in fact are the same manufacturer.

Many thanks.
David

[1] http://www.w3.org/TR/2004/REC-owl-ref-20040210/#sameAs-def

[2] http://www.w3.org/TR/2004/REC-owl-ref-20040210/#equivalentClass-def

Received on Sunday, 22 January 2006 13:13:01 UTC