[LC response] To Jeremy Carroll

Dear Jeremy,

Thank you for your comment
      <http://lists.w3.org/Archives/Public/public-owl-comments/ 
2009Jan/0051.html>
on the OWL 2 Web Ontology Language last call drafts.

We will deal with your general comments regarding motivation in a  
separate email. In this email we will address your detailed comments  
about specific aspects of the language, dealing with each of these  
separately below. Many of these comments raise specific questions  
w.r.t. motivation and use cases. Please note that New Features and  
Rationale (NF&R) [1] is not intended to provide comprehensive use  
cases for all the features of the language, but rather to provide  
examples that illustrate the rationale behind the new features of OWL  
2. As noted in its abstract NF&R document should be read in  
conjunction with the OWL Use Cases and Requirements document [2].

Regarding your specific comments:

Danger of bias: This seems to be part of your general comments  
regarding motivation which we will deal with in a separate email.

RDF interoperability: As we mentioned above, NF&R only covers new  
features. The motivation for RDF interoperability in OWL 2 is carried  
over from OWL 1 and is achieved via the same design principles-- 
nothing has changed in this respect.

Effective?: The abstract has been completely rewritten and no longer  
mentions effective reasoning algorithms.

OWLED: This again seems to be part of your general comments regarding  
motivation which we will deal with in a separate email.

Manchester Syntax: This is not a Last Call Working Draft, and the  
working group has decided that it will not be part of the  
recommendation but will be published as a working group note. It may  
be worth pointing out, however, that it is widely used, e.g., in  
TopBraid Composer, the Protege editor and the OWL 2 Primer.

OWL/XML: It should be noted that RDF/XML is the only syntax that MUST  
be supported by implementations; support for the XML syntax is not  
required (see also FH3). The XML syntax is motivated by the desire to  
support OWL users who want better interoperability with XML based  
tools and languages, for example WSDL. An additional benefit is that  
XML data can be exposed to RDF/OWL applications using GRDDL (see  
[15]). We will extend NF&R to better motivate the need for an XML  
syntax.

Links to Wiki should be links to TRs: Thank you for pointing this  
out. We will investigate the cause of the problem and ensure that it  
does not recur fix in future versions of the document(s).

Syntax examples should include RDF: This deficiency has been noted  
and the current version of the Syntax document allows readers to see  
the RDF graph version of all examples.

DisjointUnion subPropertyof UnionOf: You suggest "adding the  
assertion that owl:disjointUnionOf rdfs:subPropertyOf owl:unionOf".  
However, this doesn't make sense from a semantic point of view (see  
[17]).

DisjointUnion and DisjointClasses: As you rightly say, these  
constructs do not extend the expressive power of the language, but  
only add "syntactic sugar". However, users of OWL have pointed out  
that the lack of such constructors is very inconvenient and can lead  
to a significant blow-up in the size of the ontology if large numbers  
of pair-wise disjointness axioms need to be added -- see, e.g., [3].  
The triple form of the new n-ary axioms and its relationship to the  
binary form exactly parallels the case for differentIndividuals, so  
it should not add significantly to the implementation burden.  
Implementers within the working group, including those working on  
triple-based implementations, do not consider these features  
difficult to implement.

Negative Property Assertions: Such assertions are useful in  
applications, e.g., to state that a person does not live at a given  
location or that their age is not a given value (see [5]). Using  
complements of hasValue restrictions is cumbersome, and it is hard to  
see how this would be easier for RDF based systems to process or how  
it would improve RDF interoperability. The OWL 2 RDF Based Semantics  
includes semantic conditions that deal with negated property  
assertions [6] and thus ensure interoperability.

SelfRestriction and the Schneider variant of the Patel-Schneider  
paradox: The ability to express additional "properties of  
properties", such as reflexivity, irreflexivity and asymmetry, is  
often needed in applications -- see, e.g., [7]. SelfRestriction  
provides for "localised" reflexivity. This may be more useful in  
practice that global reflexivity. The OWL 2 RDF-Based semantics is  
not so susceptible to paradox problems, and so neither reflexivity  
nor SelfRestriction introduces any particular difficulties. NF&R  
discusses the support for this new feature.

QCRs: Thank you for your supportive comments.

Reflexive, irreflexive, asymmetric and disjoint properties: see  
above. The usefulness of these features was explicitly mentioned by  
the Health Care and Life Sciences interest group in their last call  
comment -- see [13]. The Semantic Web Deployment (SWD) Working Group  
explicitly mentioned the usefulness of property disjointness for  
specifying the semantics for SKOS mapping relations, and the  
potential usefulness of reflexivity and asymmetry -- see [14].  
Reflexivity can also be used to "approximate" SelfRestriction in  
profiles that do not support this feature. NF&R discusses the support  
for these new features.

Property chain inclusion axioms: At F2F5, the working group resolved  
to change the current RDF encoding of sub property chain axioms to a  
single triple form, which avoids the blank node on the left hand side  
and the use of "subPropertyOf" [4a]. The new vocabulary term  
"owl:propertyChainAxiom" has been introduced. Several documents were  
modified to effect this change - the best places to see the  
difference is in RDF-Based Semantics [4b], the Mapping to RDF was  
also changed [4c].

unary datatypes: Thank you for your supportive comments.

N-ary datatype: Please see [10], where it says: "This specification  
currently does not define data ranges of arity more than one;  
however, by allowing for n-ary data ranges, the syntax of OWL 2  
provides a "hook" allowing implementations to introduce extensions  
such as comparisons and arithmetic." I.e., n-ary datatypes are not  
supported, but the language is designed so as to facilitate future  
extensions in this direction. Such an extension will be published as  
a working group note (currently under preparation -- see [16]).

Annotations: We are glad to hear your general welcome for the  
improvements in the annotation system. Regarding the use of  
reification in annotations, this is mainly done in order to provide  
for annotations on axioms, a crucial feature in many tools and  
applications, as is suggested in your comment (note that this  
requirement was identified in the OWL Use Cases and Requirements  
document [2]). This is necessary, for example, on a simple subClassOf  
axiom relating two named classes; in this case there is no  
"convenient blank node", so your proposed solution does not work. In  
other cases, however, OWL 2 does attach annotations to blank nodes.  
RDF tools are accommodated in the existing design by including both  
"reified" and standard triple encodings of annotated axioms. The  
decision not to use RDF reification properties was taken in order to  
avoid overloading the meaning of the RDF vocabulary, and in response  
to comments from the RDF community (see Issue-67 [9]).

Profiles: Thank you for your generally supportive comments.

Appendix and dependency on HCLS: As mentioned above, NF&R is only  
intended to provide examples that illustrate the rationale behind the  
new features of OWL 2. The HCLS community has been very energetic and  
has contributed many of these examples. We would welcome any examples  
that you might like to contribute from your user base to broaden the  
base of use cases in NF&R.

UC#10 and UC#11: UC#10 and UC#11 motivate a feature which the working  
group was not able to fully develop, but for which it intends to  
publish a note [16]. It is true that these motivate a feature that  
did not ultimately become part of the language, although as we have  
discussed above the language does include the relevant "hooks" and an  
n-ary datatype extension will be published as a WG note [16].  
Moreover, documenting the motivation for unimplemented features may  
be of use to future working groups and is consistent with the  
approach taken in the OWL Use Cases and Requirements document [2].


TopBraid Composer: This has now been corrected.

Trademark: This has now been corrected.

Error in Appendix and Informative?: We have deleted the reference to  
TBC in the IETF application. However, although support for Manchester  
syntax is optional, the working group still considers it worthwhile  
to continue with the mimetype registration.


GRDDL: The working group has resolved to add GRDDL support to the OWL  
XML syntax (see [15]).

Informative?: See the above comment on OWL/XML. Although support for  
the XML syntax is optional, the working group still considers it  
worthwhile to continue with the MIME type registration. MIME type  
registration is not related to normativity.


XSD datatypes: Some aspects of the OWL 2 use of XSD datatypes have  
been changed, in particular OWL 2 now makes the value spaces of all  
XSD primitive datatypes pairwise disjoint, as in XSD. Because of this  
change, owl:realPlus has been removed from OWL 2.

owl:real: The new numeric datatypes specific to OWL 2 have been added  
partly to support reasoning with n-ary datatypes [16]. Unions of  
other datatypes are not adequate for this purpose.

owl:datetime: This has been superseded by xsd:dateTimeStamp (see [12]).


[1] http://www.w3.org/2007/OWL/wiki/New_Features_and_Rationale
[2] http://www.w3.org/TR/webont-req/
[3] http://lists.w3.org/Archives/Public/public-owl-wg/2007Nov/0250.html
[4a] http://www.w3.org/2007/OWL/meeting/2009-02-24#resolution_6
[4b] http://www.w3.org/2007/OWL/wiki/index.php?title=RDF- 
Based_Semantics&diff=18494&oldid=18493
[4c] http://www.w3.org/2007/OWL/wiki/Mapping_to_RDF_Graphs
[5] http://www.w3.org/2007/OWL/wiki/ 
New_Features_and_Rationale#F3:_NegativeObjectPropertyAssertion_NegativeD 
ataPropertyAssertion
[6] http://www.w3.org/2007/OWL/wiki/RDF- 
Based_Semantics#Semantic_Conditions_for_Negative_Property_Assertions
[7] http://www.w3.org/2007/OWL/wiki/ 
New_Features_and_Rationale#F6:_Reflexive.2C_Irreflexive.2C_Asymmetric
[8] http://www.w3.org/TR/owl-xmlsyntax/
[9] http://www.w3.org/2007/OWL/tracker/issues/67
[10] http://www.w3.org/2007/OWL/wiki/Syntax#Data_Ranges
[11] http://www.w3.org/2007/OWL/tracker/issues/97
[12] http://www.w3.org/TR/xmlschema11-2/
[13] http://lists.w3.org/Archives/Public/public-owl-comments/2009Jan/ 
0027.html
[14] http://lists.w3.org/Archives/Public/public-swd-wg/2009Jan/0084.html
[15] http://www.w3.org/2008/06/wiki_scribe/?source=http://www.w3.org/ 
2007/OWL/wiki/Chatlog_2009-02-24#resolution_16
[16] http://www.w3.org/2007/OWL/wiki/ 
Data_Range_Extension:_Linear_Equations
[17] http://www.w3.org/2007/OWL/wiki/ 
Direct_Semantics#Class_Expression_Axioms


Please acknowledge receipt of this email to <mailto:public-owl- 
comments@w3.org> (replying to this email should suffice). In your  
acknowledgment please let us know whether or not you are satisfied  
with the working group's response to your comment.

Regards,
Ian Horrocks
on behalf of the W3C OWL Working Group

Received on Thursday, 19 March 2009 12:12:23 UTC