- From: Barclay, Daniel <daniel@fgm.com>
- Date: Wed, 12 Aug 2009 13:39:43 -0400
- To: <public-owl-comments@w3.org>
- Message-ID: <4A82FE5F.7030208@fgm.com>
Here is some editorial feedback on the OWL 2 New Features
working draft currently at http://www.w3.org/TR/owl2-new-features/
(http://www.w3.org/TR/2009/WD-owl2-new-features-20090611/) :
* Section 2.1.1 says:
A :Lobe is exclusively either a :FrontalLobe, :ParietalLobe,
:TemporalLobe, :OccipitalLobe or a :LimbicLobe and cannot be both
of them.
- "Both" doesn't make sense for more than two things. Should be
something like "... cannot be more than one of them."
- This list is not parallel: two items have "a" and three do not.
* Section 2.1.2 says:
... it does not provide a construct for directly asserting
values of a property that an individual does not have
(negative facts).
- Is ambiguous or at least hard to understand--sounds like
referring to a _property_ that an individual does not have
(vs. to property _values_ that the individual does not have).
* Section 2.1.2 says:
NegativeObjectPropertyAssertion (resp.
NegativeDataPropertyAssertion) states that a given property
does not hold for the given individuals.
- "Respectively" doesn't apply. Should be something like
NegativeObjectPropertyAssertion and NegativeDataPropertyAssertion)
state ..."
* Section 2.1.2 says:
... (e.g. Left lung vs Right lung)
- The "e.g." should have a comma after it.
- * There are another 11 occurrences of that problem. (Try searching
for the stringfor "e.g. ".)
- The "vs" should have a period (full stop) after it.
* Section 2.2.1 says:
... processes that auto-:regulate themselves.
- Should be "... auto-:regulate ..."
* Section 2.2.2 says:
ObjectMinCardinality, ObjectMaxCardinality, and ObjectExactCardinality
(DataMinCardinality, DataMaxCardinality, and DataExactCardinality)
allow ...
- The parentheses don't make sense. Should be something like:
ObjectMinCardinality, ObjectMaxCardinality, ObjectExactCardinality,
DataMinCardinality, DataMaxCardinality, and DataExactCardinality ..
or:
ObjectMinCardinality, ObjectMaxCardinality, and
ObjectExactCardinality, and DataMinCardinality, DataMaxCardinality,
and DataExactCardinality ...
* Section 2.2.2 says:
... non negative ...
- That should be "... non-negative ..." ("non" is a prefix, not a word).
- * There are five other occurrences of that problem.
* Section 2.2.2 says:
... where n is a non negative integer, OPE an object property
expression, and [ CE ] 0 or one class expression.
- That would be clearer, especially around the "and [ CE ] 0 or one"
part, if the verb "is" were retained, as in:
... where n is a non negative integer, OPE is an object property
expression, and [ CE ] is 0 or one class expression.
- There roughly two dozen other occurrences of that pattern.
- It would also be clearer if the zero were written as the word
zero (as it is in, e.g., section 2.1.3)
- * There about about 15 other occurrences of that pattern.
* Section 2.2.2 says:
... it has exactly one direct part of each specific type frontal,
parietal, temporal, occipital, limbic lobe ...
- Some punctuation (or words) seem to be missing).
* Section 2.2.3 says:
While OWL 1 allows to assert that ...
- Should be "While OWL 1 allows assert that ..." or
"While OWL 1 allows one to assert that ..."
Section 2.2.3 says:
... - that is the property holds for all the individuals.
- The "that is" should have a comma after it.
- * There are four other occurrences of that problem.
* Section 2.2.3 says:
Everything has the same blood group as himself.
That should be either:
Everything has the same blood group as itself.
or:
Everyone has the same blood group as himself.
* Section 2.2.3 says:
Note: there are ...
- The word after "Note: " should probably be capitalized
(as it is in other notes in the document).
- * There is another occurrence in the same section.
* Section 2.2.3 says:
... asymmetry e.g., for ...
- The abbreviation "e.g." needs to have a comma before it (or
be after other appropriate punctuation).
- * There are many other occurrence of that pattern. If you
can perform a regular expression search, the regular expression
"[a-z] e\.g\." (or "^, e\.g\.") should find them.
I'm going to have to abbreviate the rest of these comments. Here's
the key to the format below:
- top-level items:
- "**" indicates something that seems to be an error (typo, grammar
problem, usage error, etc.)
- "*" indicates something that probably could be significantly
clearer (but is not an error)
- "*?" indicates something that looks wrong, but which might be
correct
- the following text is (well should be) what the document text says
- nested items:
- "-" gives a brief hint at what the (or a) problem is
- when it is just quoted text, it points out the problem or
gives replacement text
** "Many applications, which dscribe complex structures"
- "dscribe"
** "(e.g., [OBO] [RO])."
- should say something in the text, not just in the brackets;
for example, elsewhere the document says "in UML [UML]"
** "reasoning (see for example [FMA])."
- same type of problem as previous item
** the theory of [EL++] [EL++ Update].
- same type of problem as previous items
** "interpretations: for example it should be"
- "... for example, it ..."
** "allowing to state"
- should be "allowing one to state" or "allowing stating"
** "by both two different properties of the set"
- "both two"
* "contiguousTo"
- If that is a made-up example, "contiguousWith" would probably be
better. (I've only heard "contiguous with" in English, not
"contiguous to.")
** "like uncle could be defined"
- "as uncle could be defined"
** "defined, hence, it is"
- "defined; hence, ..."
** (e.g.; is:locatedIn) along another property (e.g.; partOf)
- "e.g.;" should be "e.g.,"
- "is:locatedIn" should be "isLocatedIn", right?
** "Each registered patient [...], is uniquely"
- "Each registered patient [...] is uniquely"
(Commas don't belong subject and the verb (unless they are part
of subject phrase or part of the verb phrase).)
** "For example, one could state that every person has an age which is an
integer but not to restrain the range of that datatype to say that
adults have an age greater than 18."
- "one could states ... but not to restrain" should be
"one could states ... but restrain"
- "an age which is an integer ..." should be "an age that is an
integer ..." or "an age, which is an integer, ..."
** <bullet> Extra datatypes
- The "Extra datatypes" seems to be intended to be a section heading
effectively "over" the following paragraph, but it is actually
formatted as a bullet, making in nested under the previous paragraph.
** <bullet> Datatype Restrictions
- ditto
* "DatatypeRestriction makes it also possible to ... "
- probably "DatatypeRestriction also makes it possible to ... "
** "In OWL 1 it is not possible to represent relationships between
values for one object, e.g., a square is a rectangle whose length
equals its width."
- probably " to represent relationships between values for one object,
e.g., to represent that a square is a rectangle whose length equals
its width."
** "The Data Range Extension: Linear Equations Note proposes ..."
- probably "The Data Range Extension: Linear Equations note proposes
..."
** "allows to define"
- "allows defining" or "allows on to define"
- two occurrences
** "to define a new class by a class description"
- "to define a new class with a class description"
** "explicitely"
- "explicitly"
- two occurrences
** "For ease of writing, reading, maintaining ontologies, OWL 2 provides"
- "For ease of writing, reading, and maintaining ontologies, ..."
* "datatypes that occur a couple of times in an ontology"
- better, something like "datatypes that are used multiple times in
an ontology"
** "However OWL 2 DL still imposes certain restrictions: "
- "However, OWL 2 DL still imposes ..."
(Without the comma, is sounds like the meaning of "however" in
"However you represent it, I'll read it.)
* In the HTML tables in the examples (especially see
SubAnnotationPropertyOf), the cell contents are not top-aligned,
letting the text sometime be rendered like this (view this message
in a fixed-width font):
* HCLS
The property :narrow
synonym is a subproperty
of :synonym.
OBO ontologies, in
SubAnnotationPropertyOf (:narrow_synonym particular Gene Ontology,
:synonym ) (UC#5) distinguish different kinds
of synonyms:
exact_synonym,
narrow_synonym,
broad_synonym.
- That alignment is confusing. Use "vertical-align: top" (I think)
- "distinguish different kinds" should probably be "distinguish between
different kinds"
** "allows extralogical information, such as a label or a comment, to each
ontology entity"
- something missing ("add"?)
** "OWL 2 allows to annotate ... "
- "... allows annotating ..." (or "... allows one to annotate ...")
* "provides the construct AnnotationAssertion for annotations of ontology
entities"
- probably should be "annotation of ontology entities" or
"annotations on ontology entities"
** "The IRI ... is annotated with the rdfs:label annotation property by
the human-readable label ..."
- something around "by" is wrong; as written, it says that the
label annotated the IRI with the property.
- should be something like:
"The IRI ... is annotated with the human-readable label ...
using the rdfs:label annotation property"
** "The IRI FMA:Heart of the FMA is annotated with the annotation property
FMA:UWDAID by the integer 7088 (its FMA Id)."
- similar
*? "Middle lobe of lungs are necessary right lobe since left lung do not
have middle lobe."
- What that supposed to be this?:
Middle lobes of lungs are necessarily right lungs since left
lungs do not have middle lobes.
*? "These special axioms have no semantic meaning in the OWL 2 Direct
Semantics, but carry the standard RDF semantics in the RDF-based
Semantics (via their mapping to RDF vocabulary)."
- "their mapping" or "the mapping"?
** "OBO ontologies, in particular Gene Ontology, distinguish"
- "... in particular the Gene Ontology, ..."
** "associates an entity category (class, datatype, object property, data
property, annotation property, or individual) to"
- "associates an entity category ... with"
* "OWL 2 provides in addition top and bottom object and data properties"
- probably clearer as:
OWL 2 aso provides top and bottom object and data properties ...
or:
In addition, OWL 2 provides top and bottom object and data
properties ...
** "ASCII only included"
- "ASCII only includes" (ASCII still exists, so present tense
would seem to be better)
** "This axiom does not contain an individual name for the address, so
the introduced individual is an anonymous individual."
- there is no address (e.g., street name and number)
** "This ease of writing ontologies spares from having to name an
inverse."
- "spares" doesn't seem to make sense; "ease ... comes from"
would work
* "this expression represents the inverse property of :partOf"
- probably clear as "represents :hasPart"
** "concerned by"
- should be "concerned with" ("concerned by" can be correct, but
has a different meaning)
- three occurrences
* "Other applications feel concerned ..."
- probably better as "Other applications are concerned ..."
** "e.g.; the FMA, NCI Thesaurus, SNOMED CT, Gene Ontology or some
OBO ontologies"
- probably the "or" should be "and"
* The last paragraph in section 3.1 seems out of place. It introduces
OWL 2 profiles, but they were already introduced in the preceding
paragraphs.
* "syntactical"
- could be simply "syntactic"
- several occurrences
*? "existential quantification to a class expression or a data range"
- Is it really "quantification to" or is it "quantification of"?
- several occurrences
** "Missing features include ... class negation and many other features,
see the precise list of missing features."
- "... features; see ..."
** "... too much expressive power OWL 2 EL is a language ... "
- needs period: "... power. OWL ..."
** "OWL 2 EL enables efficient implementations , E.g., CEL [CEL] is the
first reasoner for ..."
- space before comma
- "E.g." -> "e.g."
** "and (most of) expressive power of"
- "and (most of) the expressive power of"
** "etc.; Missing features are"
- "etc. Missing features are"
** "asymmetric properties, keys, see"
- "asymmetric properties, keys; see"
** "OWL 2 QL supports the same class axioms as in the structural
specification [OWL 2 Specification], except DisjointUnion which is
disallowed;"
- "disallowed."
- "except DisjointUnion, which is disallowed
** "This profile his suited"
- "... is suited"
** "on top of standard relational database"
- "on top of standard relational databases"
** "the NCI thesaurus;"
- "the NCI thesaurus."
** "existential quantification to a class, which often occurs in Life
Sciences ontologies (e.g., SNOMED) or union of class expressions
(ObjectUnionOf) are not allowed"
- needs comma before "are not allowed" (to pair with comma before
"which"
** "These restrictions prevent from infering the existence of non named
individuals which makes it possible to implement reasoners using
rule-based inference engines...."
- "prevent inferring"
- "non-named"
- probably "... individuals, which ..."
** rule extended DBMS e.g., SQL
- "rule-extended DBMS"
- "... DBMS, e.g., ..."
- multiple occurrences
** "triples, E.g., Oracle"
- "e.g.,"
** "the priority to reasoning on classes or data"
- "... priority of reasoning ... " (or "... priority given to reasoning
... "_
** "scalability etc"
- "scalability, etc.,"
** "For instance, those who look for: ... reasoning, may prefer ...;
... task, may prefer ...; ... task, may prefer ..."
- shouldn't have commas between subject and verb
** "is the RDF/XML Syntax [RDF/XML] which is the only syntax ..."
- "... Syntax [RDF/XML], which ..."
** "the Functional Syntax main purpose"
- "the Functional Syntax's main purpose"
** "motivated by the desire of better interoperability"
- "desire for"
** "all the syntaxes of OWL 2, normative and non normative."
- "non-normative"
** "first order logic"
- "first-order logic"
** "Manchestersyntax"
- "Manchester syntax"
** "Concerning the usability of the abstract syntax in OWL 2, if used
as an exchange syntax then, OWL 1 ontologies written in AS may be
input ..."
- ontologies are not used as an exchange syntax (which is what
"if used as an exchange syntax then, OWL 1 ontologies ... may be
input" says); should say something like "... if it is used ..."
*? "called XML_Serialization OWL/XML"
- doesn't look right, but I can't tell what was meant
** "schema aware editors"
- "schema-aware editors"
** "the XML savvy user"
- "the XML-savvy user"
** "blocks of OWL 2 were already present in OWL 1, albeit possibly under
a different name."
- should be "... under different names"
** "the set of inferences are the same."
- "the set ... is ..."
** "... numbers F1 to F15. (the choice done aims at conciliating an easy
understandable illustration
- "... (The ..."
- "the choice done" doesn't make sense; perhap something like "the
chosen arrangement"?
- "easily"
- "conciliating" doesn't seem right (It _sounds_ like a mistranslation
of a thought into English, but it could be that I'm not familar with
that use of the word.)
** ":ParietalLob "
- ":ParietalLobe "
** A few phrases are quoted with single quotes, although most are quoted
with double quotes.
** "Anatomy, together with Gene and Disease reference ontologies constitute
..."
- needs be either pattern "A, together with B, is ..." or
"A and B are ..." (or something equivalent); "together with" doesn't
make a plural subject, and there needs to be either zero or two
commas (not just one)
- "Anatomy" probably should be "The anatomy ontology" or "FMA"
** "upper level ontology"
- "upper-level ontology"
** "A new OWL language feature such as qualified cardinality
restriction, would be"
- Commas don't belong subject and the verb (unless they are part
of subject phrase or part of the verb phrase).
** "reflexixe"
- "reflexive"
** "The Association Class, e.g., the association between classes Person
and Company allows"
- needs comma: "Company, allows"
* "The document is not expected to change significantly, going forward."
- the comma seems unnecessary
** "Readers may selectively Show or Hide the Examples and the Functional
Syntax (FSS) or the RDF Syntax in the Examples"
- "Readers may selectively show or hide ..." (at that point, "show"
and "hide" are plain old words, not button/link labels (where
capitalization is normal))
* "Since all these constructs are simply shorthands they do not ... "
- probably better with comma: "... shorthands, they ..."
*(?) ":boundedTo :Hydrogen ... Class of objects bounded to ..."
- "bounded to" doesn't make sense; is ":boundedTo" correct, or should
it be ":bound" or ":bondedTo"?
(Please excuse any uncorrected errors in my error corrections.)
Daniel
--
(Plain text sometimes corrupted to HTML "courtesy" of Microsoft Exchange.) [F]
Received on Wednesday, 12 August 2009 17:39:48 UTC