comments on "OWL 2 Web Ontology Language Primer"

Overall it's very good, and I learned a lot. I'm suggesting potential 
areas for improvement here, most of which are just picky copy-editing 
things that will count as editorial errata. Sorry that it's coming so 
long after the Recommendation, but I figured that it was better late 
than never.

thanks,

Bob DuCharme

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

In the Abstract: "...should be read before other OWL 2 documents" would 
be better off referring to "other documents describing OWL 2" because 
section 8.2 provides a very specific meaning for the term "OWL document" 
that doesn't fit with the wording here.

Section 4.3 word usage problem "it can be excluded that there is an 
individual" (did it mean "it can be inferred"?)

4.8 word usage problem "and via their relatedness to other individuals" 
(and via their relationships)

5.2 word usage problem "are allowed to by kind of self-referential" If 
this means "be kind of self-referential," that's a little better, but 
don't say "kind of"--either it's self-referential or it's not.

5.2 word usage problem "hasChild is class member of any class" hasChild 
is *a* class member of any class?

5.3 word usage problem "we can say something about all respectively at 
least one of somebody's children"

8.2 "and also can use XML entities for namespaces." To be correct, it's 
actually using entity references, not entities, for namespace prefixes, 
and it's important to be correct here because it looks like "entity" has 
a very specific meaning in OWL 2 (e.g. in the section right after this 
one, "8.3 Entity Declarations") that has nothing to do with the concept 
of XML entities. In fact, the use of XML entity references in this 
document is confusing and adds nothing to the document or to the sample 
ontology. I realize that they're used to provide flexibility in which 
prefix represents which URL,  but doing so in this document sacrifices 
some simplicity that a Primer should not sacrifice. It would be better 
to remove them all and just say rdfs: instead of &rdfs; (etc.) throughout.

10 word usage problem "By and large, different profiles can be 
distinguished syntactically with there being inclusion relations between 
various profiles." (with the use of inclusion relations between various 
profiles?)

Received on Monday, 25 January 2010 20:44:50 UTC