[Top] [All Lists]

[ontac-forum] Adequate ontologies and better ontological analysis for en

To: <ontac-forum@xxxxxxxxxxxxxx>
From: "Gary Berg-Cross" <gary.berg-cross@xxxxxxxx>
Date: Tue, 25 Oct 2005 13:28:12 -0400
Message-id: <330E3C69AFABAE45BD91B28F80BE32C9056254@xxxxxxxxxxxxxx>

There are many things one could say in response to recent comments by Barry Smith (and Joyce Koeneman). Some of it, such as the confusion between concepts and terms goes back to my earlier, follow up question of Oct. 6 on the confusion between Ontology, knowledge, and language.  However, I think that the comments also point out one of our challenges is constructing a useful common ontology.  Much of what passes for ontology, such as those growing out of enterprise architectures (EA) are somewhat informal, even ad hoc, lacking in the level of ontological analysis we need to be useful across a wide set of domains.  I say this partially based on my experience working on enterprise architectures - I’m currently part of a team advising DoD on how to add semantics to its business enterprise architecture.  The “formalisms” for EAs are just so lacking that one winds up with ad hoc design decisions/ “fixes” to make things work. EAs and their products, for example, typically have trouble showing the integration of continuants and occurrents.  Viewing EA products we model continuant, while EA processes are occurrents, but these high-level concepts are typically not found in EAs. Instead separate models for process and product are found.

 EAs tend to look to me like logical models without a good conceptual basis.  What they have added, over time, are categories such as Roy described, but these don’t always seem rationalized and tested for ability to support the types of integrative assertions we want in an enterprise ontology.  The choice of categories and their fundamental relations is a key for us and I’m not sure it can be done in one simple step. Healthcare models, which I’ve been involved in for 15 years has done as much as anyone to come up with general categories thaqt fit their “enterprise”.  But as Barry pointed out the HL7 model, certainly not the weakest effort around, came up with a high level schema that confuses Acts and Entities so that an Observation is a sub-type of Act although Health Chart, which an Observation would be record in is an Entity.

Guarino put it well  (1998) that sometimes, "the term `ontology' is just a fancy name denoting the result of familiar activities like conceptual analysis and domain modeling, carried out by means of standard methodologies."  In 1992 John Sowa persuaded John Zachman  to co-write "Extending and Formalizing the Framework for Information Systems Architecture." JF Sowa and JA Zachman. (IBM Systems Journal, vol. 31, no. 3, 1992.), where the semantic weaknesses of information models, the roots of EA models were described, but the bulk of the EA field seems unfamiliar with this.

Sorry to be at the Ontology 101 level, but a  long time ago (1998)\ Gruber, arguing that we need objective criteria founded on the purpose of ontological model,  drafted some guidelines  to evaluate ontological designs (other than “ a priori” notions of naturalness or Truth). Below are 4 from what I believe were his preliminary set of design criteria for ontologies for knowledge sharing and interoperation.  I’m not sure that I agree with the 4th but it might be interesting to see what the group thinks about these…

1. Clarity. An ontology should effectively communicate the intended meaning of defined terms. Definitions should be objective.... Wherever possible, a complete definition (a predicate defined by necessary and sufficient conditions) is preferred over a partial definition (defined by only necessary or sufficient conditions). I take some of Barry’s comments to show that Roy’s categories are not complete or at least subject to alternative interpretations.

2. Coherence. An ontology should be coherent: that is, it should sanction inferences that are consistent with the definitions.... If a sentence that can be inferred from the axioms contradicts a definition or example given informally, then the ontology is incoherent.

3. Extendibility An ontologiest should be able to define new terms for special uses based on the existing vocabulary, in a way that does not require the revision of the existing definitions.  (we need to look ahead to integrations that will be needed, a particular problem for a general ontology)

4. Minimal ontological commitment....(Perhaps part of what the lattice discussion has been about…I’m not sure that Barry would agree with this and it might be interesting to here sides of the argument). An ontology should make as few claims as possible about the world being modeled, allowing the parties committed to the ontology freedom to specialize and instantiate the ontology as needed."

 Gary Berg-Cross

Message Archives: http://colab.cim3.net/forum/ontac-forum/
To Post: mailto:ontac-forum@xxxxxxxxxxxxxx
Shared Files: http://colab.cim3.net/file/work/SICoP/ontac/
Community Wiki: 
http://colab.cim3.net/cgi-bin/wiki.pl?SICoP/OntologyTaxonomyCoordinatingWG    (01)
<Prev in Thread] Current Thread [Next in Thread>