Review of NF&R (Action 321)

Hi Christine, Evan and all,

This review reflects the draft version of the NF&R posted at 13:40 April 
7 2009 according to the working group web site, and which presumably 
reflects incorporation of feedback to that point.

Each time I've read this document, it appears to be much improved.  I 
agreed with a number of the comments both Peter and Alan made, and 
appreciate that many of Peter's comments have been incorporated in this 
latest version.  I also agree with Alan with respect to the bullet 
headings (e.g., HCLS) in the examples, though I don't mind the boxes.

I found some additional language related issues and have a few comments 
on the structure of the document, where there appear to be some "extra" 
paragraphs to retain parallel document structure that may not be 
needed.  I've also attempted to make suggestions that might tighten the 
text a bit further here and there.  I've only addressed through section 
2, but think these comments are complementary to Alan's and Peter's 
rather than restating what they've already done.  I'll follow up with 
additional feedback on the remaining sections over the next couple of 
days, given travel constraints. 

Apologies for not being on the call today to support the discussion - 
I'm giving a talk at a conference in Tampa later today that conflicts 
with the call.  I will participate next week, however, if the discussion 
rolls over to that call.

Thanks,

Elisa

Detailed comments:

1. I prefer Peter's abstract to the one currently in the document, and 
quote it below to limit "hunting" for the text in his email:

> 0/ Abstract
>
>   OWL 2 provides new features over the previous version of OWL.  This
>   document provides a description of these new features and other design
>   choices that went into OWL 2, along with their rationale, based on use
>   cases provided to the W3C OWL Working Group.
2. Introduction
- replace  "experience of user and tool-developer" with "user and 
tool-developer experience"
- replace "an appendix" with "the appendix provided in section 8"
- paragraph breaks included in Peter's original email, which help from a 
readability perspective, appear to have been lost

3. Towards the end of the beginning of section 2, replace "Depending on 
his profile the reader can /Hide and Show/ the Examples." with "Readers 
may selectively Show or Hide the Examples by toggling the button, below."

4. Section 2.1 Syntactic sugar
- section 2.1.1: replace "It is a shorthand for separate axioms making 
the classes pairwise disjoint and one setting up the union class." with 
"It is a shorthand for owl:disjointWith statements used in combination 
with owl:unionOf to define a complete superclass from a set of mutually 
disjoint subclasses." (Peter's suggestion, which provides a more precise 
explanation)
- I think the number of examples here can be shortened to one or two - 
the additional examples don't add useful information ... you might keep 
the Lobe example only, but delete "...and cannot be both of them"
- section 2.1.2; replace "While OWL 1 provides means to state that two 
subclasses are disjoint, stating that several subclasses are pairwise 
disjoint cannot be done concisely." with "While OWL 1 provides a way to 
say that two classes are disjoint from one another, stating that several 
classes are pairwise disjoint cannot be done concisely."
- delete "It is a shorthand for binary disjointness axioms between the 
classes." (not needed)
- I agree with Alan's comment on the note
- section 2.1.3 -- I generally agree with Alan's comments here, and 
would suggest rewording anywhere the document says that something is 
"impossible" in OWL 1.  I've tried to make suggestions for some of 
these, but did not do an exhaustive search.

5. Section 2.2 New constructs for Properties
- section 2.2.2 Qualified object and data cardinality: replace 
"Qualified cardinality restrictions are present in SROIQ and have been 
successfully implemented." with "Qualified cardinality restrictions are 
supported by SROIQ." (the rest of the sentence is not useful unless you 
can also say which reasoners have implemented this).
- in the example, delete "The following examples are some examples of 
Object Property Cardinality Restrictions from Use Cases among many in 
HCLS. " (not needed)
- section 2.2.3 Reflexive, Irreflexive, and Asymmetric Object 
Properties: replace "While OWL 1 allows to assert that an object 
property is symmetric or transitive, it is impossible to assert that the 
property is reflexive, irreflexive or asymmetric." with "While OWL 1 
allows one to assert that an object property is symmetric or transitive, 
it does not provide the means to state that a property is reflexive, 
irreflexive, or asymmetric."
- in the example for ReflexiveObjectProperty, I'm not sure you need the 
note regarding the distinction among interpretations of part-of - it 
isn't the focus; same comment for IrreflexiveObjectProperty
- you might consider shortening the note at the end of 2.2.3.  I'm not 
sure you want to delete it altogether, but you could compress this 
substantially -- I agree with Alan's revision of the SKOS part of the text
- 2.2.4 Disjoint properties: rephrase "While OWL 1 provides means to 
state the disjointness of classes, it is impossible to state that 
properties are disjoint." to say something like "While OWL 1 provides 
means to state the disjointness of classes, there is no way to make 
similar statements about properties." 
- The note in the example needs to be rephrased - I would say something 
like "Note: this example was taken from Use Case #1, which provides 
domain specific definitions of the properties in question as used in 
describing brain anatomy for surgical purposes." and leave it at that.  
I had a hard time parsing the current explanation, which isn't any 
clearer after reading the use case.
-2.2.5 Property chain inclusion
- delete the first sentence, and start with "The OWL 2 construct ..."

6. 2.3 Extended datatype capabilities
- do we need a sub paragraph here, since there is only one?  if not, 
simply insert the "F10" after 2.3 and eliminate the sub-paragraph; 
otherwise:  in 2.3.1 Extra Datatypes and Datatype Facets -- change 
"Extra" to "Additional"
- in the first paragraph, second sentence, replace "... which is an 
integer but not to restrain ..." with "...which is an integer, but not 
restrict ..."
- what happened to F11?  There are references to F11 in the table in 
section 6 ... but the section describing this feature seems to be 
missing ... unless something happened to the numbering along the way

7. 2.4 Simple metamodeling capabilities
- 2.4.1 Punning (same comment as above - if we don't need the 
sub-paragraph, insert "F12" after 2.4 and add "(Punning)" at the end)
- in the first sentence, change "...names of, e.g., classes and 
individuals." to "the names of first class entities, such as classes and 
individuals."
- at the end of the second sentence, delete "unnders"
- at the end of the fourth sentence, replace "separate, as is required 
in DL reasoners" with "independent, as required to facilitate tractable 
reasoning."

8. 2.5 Extended annotations
- 2.5.1 F13: Annotations -- (same comment as above - if we don't need 
the sub-paragraph, insert "F13" after 2.5
- rephrase "...a comment, to each ontology entity, but did not allow 
annotation of axioms, e.g., with information about who asserted an axiom 
or when." to say "a comment, with entities in an ontology. It is limited 
with respect to additional annotation facilities, for representation of 
provenance, for example."
- delete "Even annotations of annotations are possible." (not needed)

9. 2.6 Other Innovation
-Rename this section 2.6 F14 Declarations and eliminate the subparagraph 
if possible.  If not, then the primary heading should be "Additional 
Features" or something like that.

Received on Wednesday, 8 April 2009 06:04:45 UTC