W3C home > Mailing lists > Public > public-owl-wg@w3.org > April 2009

RE: Action 311-- Review of SS&FS document

From: Boris Motik <boris.motik@comlab.ox.ac.uk>
Date: Mon, 6 Apr 2009 12:36:11 +0100
To: "'Bernardo Cuenca Grau'" <Bernardo.Cuenca.Grau@comlab.ox.ac.uk>, "'OWL Working Group WG'" <public-owl-wg@w3.org>
Message-ID: <AB71485A061140F8A7158F78BED83A9B@wolf>
Hello Bernardo,

Thanks a lot for your comments. Please find my answers inline.

Regards,

	Boris

> -----Original Message-----
> From: public-owl-wg-request@w3.org [mailto:public-owl-wg-request@w3.org] On
> Behalf Of Bernardo Cuenca Grau
> Sent: 03 April 2009 11:03
> To: OWL Working Group WG
> Subject: Action 311-- Review of SS&FS document
> 
> 
> The document is currently in a great shape. My comments below are very
> minor.
> 
> 
> * Section 1 *
> 
> 
> `For example, a class expression describes a set of individuals in terms
> of the restrictions on the individuals' features'
> --> Not clear. Probably it would be better to replace the word `feature'
> by something else.
> 

I've changed it to "characteristics".

> 
> * Section 2.1 *
> 
> `Sets written in one of the linear syntaxes ' --> the terms `linear
> syntax' and `exchange syntax' are sometimes used
> intercheangeably in the document. Probably the document should refer to
> theses syntaxes in a consistent way. I would suggest
> dispensing with the term `linear syntax'.
> 

I've changed "linear syntaxes" to "exchange syntaxes". This was the only place
where "linear" was used.

> * Section 2.3 *
> 
> The link [UNICODE] seems broken. Same happens with the link [XML 1.0].
> 

All references are in the process of being updated to a common format. Once this
is done, all links will be fixed.

> * Section 3 *
> 
> 
> `The following list summarizes all the conditions that O is required to
> satisfy to be an OWL 2 DL ontology. ' -->  should
> say `The following list summarizes all the conditions that *an OWL 2
> ontology* O is required to satisfy
> to be an OWL 2 DL ontology. '
> 

Thanks!

> The last of these conditions says: `Each O' directly imported into O
> MUST satisfy all of these restrictions as well. '
> It should probably say: ``Each O' *imported* into O MUST satisfy all of
> these restrictions as well. ''
> 

I believe "directly imported" suffices, as it includes the last condition.
(Thus, the ontologies directly imported into each directly imported ontology
need to satisfy this.) In other words, the condition is recursive.

> * Section 3.2 *
> 
> `To access a particular version of OI, one needs to know that version's
> version IRI VI' --> This sentence seems broken.
> 

I'm not sure what you refer to here: the sentence seems OK to me. I've just
changed it mildly like this:

To access a particular version of ''OI'', one needs to know that version's
version IRI ''VI''; the ontology document of the version <em title="SHOULD in
RFC 2119 context" class="RFC2119">SHOULD</em> then be accessible from ''VI''.

> 
> * Section 3.4 *
> 
> The link to Section 3.6 seems broken.
> 

Thanks! I've checked all the links using a link checker and have fixed all
broken links apart from the ones to the references. As I've said before, the
links to the references will be fixed once the reference format has been agreed
upon.

> * Section 3.7 *
> 
> Reference [UNICODE] seems broken
> 
> Reference to Section 3.6 seems broken
> 
> 
> * Section 4.1 *
> 
> Description of the value space of owl:rational. The specification says
> the following:
> `It is a subset of the value space of owl:real, and it contains the
> value space of xsd:decimal (and thus
> as well of all other xsd: numeric datatypes from the above list)'.
> 
> It is not clear which exactly are the numeric datatypes in the above
> list: all of them? a subset of them?
> 

I'm not sure I understand this comment: the idea was to say that the value space
of owl:rational includes the value spaces of all xsd: datatypes from the above
list; hence, it is not a subset, but all of them. I've slightly rephrased the
sentence like this:

* The value space of ''owl:rational'' is the set of all rational numbers. It is
a subset of the value space of ''owl:real'', and it contains the value space of
''xsd:decimal'' (and thus of all ''xsd:'' numeric datatypes listed above as
well).

> * Section 4.2 *
> 
> `The data property with IRI owl:topDataProperty connects all possible
> individuals with all literals.
>  (In the DL literature this is often called the top role.) ' --> I don't
> think it is called top role. Actually, I
> don't know whether this special role has been defined in the DL
> literature. The top role often refers to the
> one that is interpreted as the universal binary relation over the
> abstract domain.
> 

To prevent any confusion, I've removed any references to the top and bottom
roles in the DL literature.

> * Section 5.6 *
> 
> `Individuals in the OWL 2 syntax represent actual objects (semantic
> individuals) from the domain' --> the
> expression `semantic individual' is confusing.
> 

Agreed: I removed any mention of it.

> * Section 5.6.2 *
> 
> Replace  `the axiom closure of an ontology O' by  `the import closure of
> an ontology O' (or vice versa). In any case, axiom closure and
> import closure seem to be used to denote the same notion in this section.
> 

Axiom closure and import closure of O are two different things: the first one
contains all axioms from O and all the (directly and indirectly) imported
ontologies, whereas the second one contains O and all (directly and indirectly)
imported ontologies. Hence, I'm afraid I cannot replace one with the other.

> ` (i.e., with an anonymous individual having a globally unique node ID).
> ' --> Not clear what `globally unique' means
> in this context.
> 

I've changed this as follows:

    ... (i.e., an anonymous individual whose node ID is unique in the import
    closure of ''O'')

> 
> * Section 5.8 *
> 
> `Each IRI I used in an OWL 2 ontology O can, and sometimes even must, be
> declared in O' --> Probably the words `can' and
> `must' should be written in italics for consistency with the rest of the
> document.
> 

This sentence was not meant to be normative; therefore, it does not use the RFC
keywords. To avoid any confusion, I've replaced "must" with "needs to be".

> * Section 5.8.1 *
> 
> 
> `If an object property with an IRI I occurs in some axiom in Ax, then I
> is declared in Ax as an object property.'
> --> It is not clear how one would know whether I is an object property
> if it has not been declared. Obviously it should
> be possible to know it from the way the IRI occurs in the ontology, but
> this probably deserves some explanation.
> Same comment applies to the remaining bullets.
> 

This section talks about Ax as a set of structures from the structural
specification. The structural specification is fully typed, so you can always
check whether an object property with IRI I occurs in an axiom in Ax. This
restriction merely says "if Ax contains an instance of the ObjectProperty UML
class with an IRI I, then Ax must contain a declaration for I".

> * Section 5.8.2 *
> 
> `Although declarations are optional for the most part,... ' --> This is
> not quite clear. Section 5.8.1 seems to imply
> that declarations should always exist in OWL 2 DL. So, it is likely that
> many people will like their ontologies to
> be in OWL 2 DL and in many cases declarations would actually be compulsory.
> 

Declarations are needed for classes and properties, but not for individuals. I
agree, however, that "for the most part" was confusing. Therefore, I've
rephrased the sentence like this:

    Although declarations are not always required, they can be used to catch
    obvious errors in ontologies.

> 
> * Section 6.2 *
> 
> `notion of data property expressions'--> replace with `notion of data
> property expression'
> 

I've removed "the notion of".

> 
> * Section 9.2.1 *
> 
> First example. `Having a dog is a kind of having a pet' --> sounds
> strange. it would probably be better to say that
> `Having a dog implies having a pet'.
> 

I was wondering how to phrase subproperties more naturally -- many thanks for
this suggestion. I've replaced similar phrases throughout the document.

> * Section 9.3.6 *
> 
> End of second example.
> `which violates the functionality restriction on a:numberOfChildren and
> makes the ontology is satisfiable' --> this is
> obviously broken. It should say something like `and makes the ontology
> unsatisfiable'.
> 
> 

Thanks!

> * Section 9.4 *
> 
> `This axiom allows one to use the defined datatype DT as a synonym DR'
> --> should say:
> `This axiom allows one to use the defined datatype DT as a synonym *of* DR'.
> 

I've added "synonym for".

> * Section 9.6 *
> 
> `The DataPropertyAssertion axiom allows one to state that an individual
> is connected by a data property expression to literal' -->
> `The DataPropertyAssertion axiom allows one to state that an individual
> is connected by a data property expression to *a* literal'
> 

Thanks!

> * Section 9.6.6*
> 
> In the example, replace ` By the second axioms' with  `By the second axiom'
> 

Thanks!

> * Section 10.2 *
> 
> `These statements are treated as axioms only in order to simplify the
> structural specification of OWL 2.' --> Probably add another
> explanatory sentence saying that annotations, unlike other axioms, are
> not treated as logical statements.
> 

This is the case only in Direct Semantics, but not in the RDF-Based Semantics. I
believe that "less is more" approach should be adopted here: we already had
debates about this problem.

> * Section 11.2 *
> 
> In the example, replace `The second axioms depends on the first one'
> with `The second axiom depends on the first one'.
> 

Thanks!
Received on Monday, 6 April 2009 11:37:23 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 6 April 2009 11:37:24 GMT