- From: Phil Tetlow <philip.tetlow@uk.ibm.com>
- Date: Wed, 11 Jan 2006 14:01:21 +0000
- To: Jeremy Carroll <jjc@hpl.hp.com>
- Cc: public-swbp-wg@w3.org
Jeremy, Many thanks for your input. I guess we should wait for your complete review of the updated version before preceeding with your comments? Can we assume a couple of weeks for further review? Best Regards, Philip Tetlow Senior Consultant (Certified Technical Architect) IBM Business Consulting Services Mail: IBM United Kingdom Limited, 1175 Century Way, Thorpe Park, Colton, Leeds, LS15 8ZB Current Assignment: Mobile: +44 (0)7740 923328 Email: philip.tetlow@uk.ibm.com Jeremy Carroll <jjc@hpl.hp.com> To 11/01/2006 13:34 Phil Tetlow/UK/IBM@IBMGB, Holger Knublauch <holger@knublauch.com> cc public-swbp-wg@w3.org Subject Re: [SE] comments on primer hi Summary: still minor fixes required on numbers 7 and 8. I include suggested text for these two cases. Also note that the text has changed significantly since my review, and I have not reviewed the new text. Looking at: http://www.w3.org/2001/sw/BestPractices/SE/ODSD/20051217/ and picking up thread form Nov 2005. Phil Tetlow wrote: > 1. Note that David Wood suggested moving para starting "The Semantic Web > community has ..." forward so that it is clearer what the note offers. > > Agree (I think), but given that the abstract is now shorter and more > focused on Sem Web languages, I personally think that the preamble in the > introduction is acceptable. OK, no further change on this one. > > 2. It is unclear whether the abstract is appropriate. > > Suggest the abstract is shortended to: > > Domain models play a central role throughout the software development > cycle, from requirements analysis to design, through implementation and > beyond. As such, great progress has been made in the consistent use of > models throughout this process. Modern software development tools with UML > support and code generation as well as Model-Driven Architectures allow for > developers to synchronize and verify technical implementation with user > requirements using models - the ultimate deliverable of computer code still > only being another form of model at the end of the day. However, there are > still challenges which can be addressed by Semantic Web languages such as > RDF Schema and OWL, which have been optimized to build and reason with > domain models. This note is hence intended to act as an introduction to > Semantic Web technologies for software developers with background in > object-oriented languages like UML and Java. Its aim is to clarify the > differences between RDF/OWL and object-oriented languages, and to encourage > mainstream developers to add Semantic Web technology to their routine tool > kit. > Current abstract, which differs from above is OK > 3. The reference [MvH 2004] should be [OWL] or similar. > > Agree... > Fixed > 4. The reference [MM 2004] should be [RDF] or similar. > > Agree.. > Fixed > 5. Suggest rephrasing of "Semantic Web languages are far more specific." > > Perhaps change to "Semantic Web languages are far more formal and > semantically precise" > Text has changed sufficiently since I reviewed it, that this comment seems to be superceded. > 6. Suggest replace "In contrast to object-oriented languages" with > "In contrast to many object-oriented languages" > ("many" could be "well-known" or "most" or "some" or "several"; I prefer > "many") > > I agree with Jeremy... > Fixed > 7. The para stating "In order to "attach" a property to a class" is not > well-worded, suggest revising for greater clarity and accuracy > > Suggest possibly? > > In order to "attach" or "associate" a property with a class, rdfs:domain > statements can be used. rdfs:domain is a tag from the RDF Schema namespace > that relates a property to a class using predication. In the example above, > the domain of hasPrice is Product. As such, from an object-oriented point > of view this would mean that all instances of the Product class could have > price values associated with them, hence making Price a Product attribute. > However, in RDF and OWL this also has additional connotations: any resource > that is the object of hasPrice, (i.e. normally a dicreet price value), must > also relate to the subject of Product. In other words, a domain statement > in RDF can be used to classify value instances. Therefore, pointing back to > the above example, if something has a price, then it can be handled as an > instance of Product, even if it partakes in other declarations (or triples) > - a crutial matter that will be discussed in more detail later in the > context of reasoning with OWL. > > Not OK. (Sorry I should have commented on this proposed fix earlier) Current text reads: [[ In order to "attach" or "associate" a property with a class, rdfs:domain statements can be used. rdfs:domain is a tag from the RDF Schema namespace that relates a property to a class using predication. In the example above, the domain of hasPrice is Product. As such, from an object-oriented point of view this would mean that all instances of the Product class could have price values associated with them, hence making Price a Product attribute. However, in RDF and OWL this also has additional connotations: any resource that is the object of hasPrice, (i.e. normally a discrete price value), must also relate to the subject of Product. In other words, a domain statement in RDF can be used to classify value instances. Therefore, pointing back to the above example, if something has a price, then it can be handled as an instance of Product, even if it partakes in other declarations (or triples) - a crutial matter that will be discussed in more detail later in the context of reasoning with OWL. ]] ==> my suggestion [[ In order to "attach" or "associate" a property with a class, rdfs:domain statements can be used. rdfs:domain is a tag from the RDF Schema namespace that relates a property to a class using predication. In the example above, the domain of hasPrice is Product. As such, from an object-oriented point of view this would mean that all instances of the Product class could have price values associated with them, hence making Price a Product attribute. However, in RDF and OWL this also has additional connotations: any resource that is the subject of hasPrice is an instance of the Product class. In other words, a domain statement in RDF can be used to classify instances. Therefore, pointing back to the above example, if something has a price, then it can be handled as an instance of Product, even if it partakes in other declarations (or triples) - a crucial matter that will be discussed in more detail later in the context of reasoning with OWL. ]] The current text (as well as spelling on "crucial", confuses domain and range. Domain makes no comment whatsoever about the object of hte hasPrice triple. > 8. The use of RDF Individual in Figure 6 is useful but misleading, in > that RDF does not provide such a class. owl:Thing is perhaps the > closest. Perhaps the best thing is simply to leave the figure as is but > add a footnote to the effect that RDF Individual is introduced in this > note for explanatory purposes. > > Agree Not fixed. Please fix it. e.g. [[ As illustrated in Figure 4, a typical object model to represent Semantic Web ontologies would contain classes to represent resources, classes, properties and individuals. ]] ==> [[ As illustrated in Figure 4, a typical object model to represent Semantic Web ontologies would contain classes to represent resources, classes, properties and individuals. Note that the terms RDFSClass and RDFProperty relate to the classes rdfs:Class and rdf:Property defined in RDF Schema, whereas the term RDFIndividual has no direct counterpart defined in RDF Schema. ]] > > Holger...hope this helps > > > > > >
Received on Wednesday, 11 January 2006 14:02:36 UTC