W3C home > Mailing lists > Public > public-swbp-wg@w3.org > April 2006

attribute interoperability

From: Koen Aerts <koen.aerts@sadl.kuleuven.be>
Date: Tue, 11 Apr 2006 12:20:04 +0200
Message-ID: <443B82D4.50302@sadl.kuleuven.be>
To: public-swbp-wg@w3.org

Dear all,

In our project we are developing a system to gain semantic
interoperability between different topographic databases using OWL-DL.
Such databases contain topographic features with their attributes. For
example roads with attributes road width, road surface, etc.
At this moment we have reached semantic interoperability of topographic
feature classes. We can infer equivalences and sub/super class
relationships between topographic feature classes of different
databases. For example we can inter the equivalency between a 'road' in
an english topographic database and a 'route' in the french counterpart.

In the next step we are trying to match the equivalent attributes. We
want to derive that attribute 'a' of topographic feature class 'A' is
equivalent to attribute 'b' of topographic feature class 'B', given
their is a equivalency or a sub/superclass relationship between 'A' and
'B'.
 
We want to use attributes 2 ways:
1) to derive to which topographic feature class a topographic feature
belongs according to it's attribute values
2) once we know the topographic feature class of a topographic feature,
to fill in the attribute values for a specific attribute from an
equivalent attribute from a different database.

Here is my question then: how would you model this in OWL?
When we look at other projects, attributes are modelled as properties in 
OWL. The disadvantage of this approach is that you can not reason about 
properties. It is possible to define equivalent properties, but then you 
are doing the reasoning yourself. We want a reasoner to do this for us.

So we made a topographic upper ontology which can be used to define 
topographic features and their attributes. Both are modelled as 
classes.  We make seperate ontologies for each topographic database 
where the topographic features and attributes of that database are 
defined in terms of our topographic upper ontology. The genereal object 
property "hasAttribute" relates a topographic feature with it's attributes.

The main advantage of this approach is the ability to reason about the 
attributes. Two topographic databases can be described independent from 
each other, the reasoner will infer the relations between the two.

We would like to invoke a discussion on how to deal with attributes and 
the advantages and disadvantages of different approaches.

Kind regards,
Koen

-- 
Koen Aerts
K.U.Leuven
R&D Division SADL (Spatial Applications Division Leuven)
Geo-Institute, Celestijnenlaan 200 E, B-3001 Heverlee, Belgium
TEL: +32-16-32.97.27 FAX: +32-16-32.97.24
http://www.sadl.kuleuven.be 


Disclaimer: http://www.kuleuven.be/cwis/email_disclaimer.htm
Received on Tuesday, 11 April 2006 12:39:55 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 17:31:19 UTC