[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: Re: [set] Requirements: versioning and corrections
Dear Stephan, OWL allows versioning but we have not yet used this in practice. We can address this issue later when we progress in our work in SET TC and face the problem. Thanks, Asuman Stephen Green wrote: > In line with my previous discussion of requirements > for versioning for the SET approach to ontological > management of contexts for document interoperability > how about we write a white paper or position paper > to propose a solution if the problem I have suggested > does in fact exist. > > I would propose something along the following lines: > > As a real example take the latest UBL 2.0 errata 01 > specification which among other things caters for > an error in the BOSNIA AND HERZEGOVINA code. > Say there is an ontology which includes the old code: > when making any inferences using this ontology > a reasoner will report a certain value for BOSNIA > but not one for BOSNIA AND HERZEGOVINA. Now > feed the reasoner with an ontology based on the > update package (errata) and run the reasoner > again. If OWL is used, the original code information > cannot just using OWL be eliminated from the > inferences. Instead the reasoner will now report > still report that BOSNIA has code 'AND' (this is the > error) and it may also have a correspondence > between BOSNIA and BOSNIA AND HERZEGOVINA > such that it reports two values for BOSNIA - the > old, erroneous one and the new, correct one. > > What I would propose as a solution is that there > be a standard to list in order the ontologies by > namespaces or other universally unique IDs in the > order they have to be entered into a completely > clean knowledgebase immediately prior to > running the reasoner after which the knowledgebase > is either frozen or cleared - when OWL is used. > This proposed standard would amount to a kind of > reasoner's knowledgebase recipe. It would ensure > a certain degree of uniformity across implementations > (though reasoners may differ in their inferences > perhaps). It would also allow for conformance testing > of implementations. One possible recipe language > might be a test assertions markup (includes > prerequisites, and formal predicates). > > > Does this ring true at all with SET TC experiences > with reasoners and knowledgebases with OWL? > > All the best > > Steve > > -- ____________________________________________________________________________ Professor Asuman Dogac email: asuman@srdc.metu.edu.tr WWW: http://www.srdc.metu.edu.tr/~asuman/ Director Phone: +90 (312) 210 5598, or Software R&D Center +90 (312) 210 2076 Department of Computer Eng. Fax: +90 (312) 210 5572 Middle East Technical University +90 (312) 210 1259 06531 Ankara Turkey skype: adogac
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]