Re: CORRECTION Terminology: Data Source & Domain Ontologies

I agree with Kingsley that we should start using specific terms in our
lexicon.

During our F2F, I used the word "putative" ontology to describe the Data
Source Ontology, that is obtained through "direct mapping". Somebody also
recommend the use of "candidate ontology". This word brought a bit of
commotion because of the philosophical definition. My use of the word
putative is to describe the quality of the Data Source Ontology. A direct
mapping system will generate straight forward data source ontology (table->
class, FK -> object property, etc). The quality of this resulting ontology
is not sufficient by itself, therefore the need to map it to an Domain
ontology. However, there may be the case that if the relational schema does
portray sufficient domain semantics, then the data source ontology can be
enough. Hence the ontology is putative because you it may be semantically
rich or not, depending on your needs.

Just wanted to make this clear, from my point of view.

Juan Sequeda, Ph.D Student

Research Assistant
Dept. of Computer Sciences
The University of Texas at Austin
http://www.cs.utexas.edu/~jsequeda <http://www.cs.utexas.edu/%7Ejsequeda>
jsequeda@cs.utexas.edu

http://www.juansequeda.com/

Semantic Web in Austin: http://juansequeda.blogspot.com/


On Fri, Nov 21, 2008 at 7:07 PM, Kingsley Idehen <kidehen@openlinksw.com>wrote:

>
> Kingsley Idehen wrote:
>
>>
>> All,
>>
>> While digesting today's minutes, I came across the terms: Data Source
>> Ontology and Domain Ontology. Although this isn't the first time I've
>> encountered these terms (since I've met Prakash at our month Semantic Web
>> Gatherings a few times), I would like to acknowledge that these terms should
>> become vital parts of the RDB2 RDF lexicon for the following reasons:
>>
>> 1. Historic links to related realms e.g. in the early '90's NeXT developed
>> an Objective-C based mapping layer called EOF (Enterprise Object Frameworks)
>> the enabled RDBMS to Objective-C mapping via what is Prakash calls a "Domain
>> Ontology" generated from the source RDBMS schema
>>
>>  I meant: "Data Source Ontology" :-)
>
>> 2.  Contemporary links to related realms e.g., Microsoft offers Enterprise
>> Object Frameworks as its .NET based mechanism for RDBMS to .NET based Entity
>> mapping via an Entity Model Generator (EDM) that also produces a "Domain
>> Ontology".
>>
>>  Ditto.
>
>>
>> Prakash/Mike: Thanks for the introduction of these terms. I believe they
>> will greatly aid the messaging aspect of the RDB2RDF effort, amongst other
>> things :-)
>>
>
> The "Data Source Ontology" is where NeXT and more recently Entity
> Frameworks have focused their attention. RDF on the other hand, extends this
> concept by adding the ability to go beyond the "Data Source Ontology"
> directly from the RDBMS or by mapping the "Data Source Ontology" to a
> "Domain Ontology".
>
> Phew! Got that sorted :-)
>
>
>>
>> Links:
>>
>> 1. http://en.wikipedia.org/wiki/Enterprise_Objects_Framework
>> 2. http://en.wikipedia.org/wiki/ADO.NET_Entity_Framework
>>
>>
>>
>
> --
>
>
> Regards,
>
> Kingsley Idehen       Weblog: http://www.openlinksw.com/blog/~kidehen<http://www.openlinksw.com/blog/%7Ekidehen>
> President & CEO OpenLink Software     Web: http://www.openlinksw.com
>
>
>
>
>
>