From:
higgins-dev-bounces@xxxxxxxxxxx [mailto:higgins-dev-bounces@xxxxxxxxxxx] On Behalf Of Anthony Nadalin
Sent: Thursday, February 01, 2007
4:59 PM
To: Higgins
(Trust Framework) Project developer discussions
Cc: 'Higgins
(Trust Framework) Project developer discussions';
higgins-dev-bounces@xxxxxxxxxxx
Subject: RE: [higgins-dev] A
simple firstname attribute in the Higgins
dialectof OWL
>In this issue, it seems to me that the display name is
part of the schema
Part of what schema ? the provider is only going to reflect the RDF that
matches the repository
Anthony Nadalin | Work 512.838.0085 | Cell 512.289.4122
"Dale Olds" <dolds@xxxxxxxxxx>
|
"Dale
Olds" <dolds@xxxxxxxxxx>
Sent by:
higgins-dev-bounces@xxxxxxxxxxx
02/01/2007 12:56 PM
|
Please respond to
"Higgins \(Trust
Framework\) Project developer discussions"
<higgins-dev@xxxxxxxxxxx>
|
|
|
To
|
"'Higgins (Trust Framework) Project developer
discussions'" <higgins-dev@xxxxxxxxxxx>
|
|
cc
|
|
|
Subject
|
RE: [higgins-dev]
A simple firstname attribute in the Higgins
dialect of OWL
|
|
I
feel I must reply since I am an example in this thread. Yes, Tony and Paul are
correct, I am so old that I have used punch cards and teletypes. It was shortly
after the discovery of fire. Now to the real issue...
BTW, I think it's GREAT that Paul is putting
together examples like this. I know I glaze over on too much OWL discussion. A
specific example helps a lot.
In this issue, it seems to me that the display
name is part of the schema -- the attribute definition. Whereas the
"real" data is actually the attribute value -- the value of and
instance of the attribute. Not the display name of the attribute type. So I am
not seeing a problem here. The display name goes in the OWL for the definition
of the attribute type understood by the CP. The data comes from the punch cards
as it is read by the CP. More specifically, the display name applies to
potentially many attribute instances, from values from many punch cards. And
there may be localized display names as well.
I think this is the same as what Paul says, but I
think of it in different terms. Value or instance data vs. type or schema data.
Is there still a problem, Tony?
--Dale
>>> "Paul
Trevithick" <paul@xxxxxxxxxxxxxxxxx> 01/31/07 6:52
PM >>>
I need to make a distinction between the *real*
data (as you call it and
I*ll continue using your term) and what a Context
Provider actually presents
at the IdAS level. The CP is responsible for
normalizing this real data and
declaring the schema/ontology of that normalized
data. As one part of this
normalization process let*s consider what the CP
does for one single
attribute in the real data. It must not only make
the value of the attribute
available, but it must also declare the type (as a
URI) of the attribute.
Let*s say that this attribute*s type URI is
declared by the CP as
http://example.com/firstname.
The CP must provide a schema (in response to
*getSchema* (soon to be renamed *export*)) that
includes a section of OWL
that describes http://example.com/firstname. Within
this OWL the CP may
choose to include a display data object with a
display label sub- object
that, for the English language, has the value
*First Name*. If the CP was
written back when Dale was going to school and all
we had was punch cards,
the recently written CP might still choose to
include a display data object
with a display label sub- object so that it can be
displayed on computers
with visual displays.
_____
From: higgins- dev- bounces@xxxxxxxxxxx
[mailto:higgins-
dev- bounces@xxxxxxxxxxx] On Behalf Of Anthony Nadalin
Sent: Tuesday, January 30, 2007 10:22 PM
To: Higgins
(Trust Framework) Project developer discussions
Cc: 'Higgins
(Trust Framework) Project developer discussions';
higgins- dev- bounces@xxxxxxxxxxx
Subject: Re: [higgins- dev] A simple firstname
attribute in the
Higginsdialect of OWL
Not following all the way here, but the display
name will most likely not
reside with the "real" data as the real
data was created years ago and they
had no displays (back when Dale was going to
school), so I'm not seeing that
linkage here
Anthony Nadalin | Work 512.838.0085 | Cell
512.289.4122
Inactive hide details for "Paul Trevithick"
<paul@xxxxxxxxxxxxxxxxx>"Paul
Trevithick" <paul@xxxxxxxxxxxxxxxxx>
"Paul Trevithick"
<paul@xxxxxxxxxxxxxxxxx>
Sent by: higgins- dev- bounces@xxxxxxxxxxx
01/30/2007 11:12 AM
Please respond to
"Higgins
\(Trust Framework\) Project developer discussions"
<higgins- dev@xxxxxxxxxxx>
To
"'Higgins
\(Trust Framework\) Project developer discussions'"
<higgins- dev@xxxxxxxxxxx>
cc
Subject
[higgins- dev] A simple firstname attribute in the
Higgins dialect of OWL
I*ve been working up a simple attribute example as
we evaluate two proposed
extensions to the current higgins.owl:
(a) Valery*s DisplayData extension (very
preliminary. Valery is I think
working on a new version
(b) To support arbitrary XML Schema- defined
datatypes, I propose that we
copy the body of this ontology
http://protege.stanford.edu/plugins/owl/owl- library/2005/08/07/xsp.owl
into
higgins.owl. Note: The issue of how to use XML
Schema types within the OWL
community is still evolving. The protégé team (the
source of the above link)
has developed an approach that is suited to our
needs, so I'm proposing here
that we adopt it. We may in the future choose to
update higgins.owl to
conform to the approach that the larger OWL
community settles on.
The use of the (b) above necessitates removing the
OWL- DL restriction and
using OWL- Full in higgins.owl as we discussed a
bit at the F2F last week.
The rest of this email has been cross posted
(without the above) to
idschemas.org due to great interest there in folks
understanding how the
Higgins dialect might be used to support a number of projects
including
attribute extensions to OpenID.
The exercise is to describe these semantics about
a "first name" attribute:
o it has exactly one value of type string
o this value is restricted to be constructed of
the characters
a- z or A- Z -- a bit silly, but you
get the idea
o it has an English display string of "First
Name"
o it is semantically equivalent to
"givenName" as defined by
rivalattributes.com
Note 1: The following is based on an in- process
draft of the base
higgins.owl file (not the published one as is
implied by the URL below).
Note 2: The display data and XML Schema constructs
are new, may evolve, and
have required changing to use OWL- Full instead of
today's OWL- DL.
Note 3: Comments ## are interspersed. And to keep
the clutter down, I've
eliminated some elements e.g. version data,
labels, and comments, etc.
commonly used by developers.
- Paul
## We start off with the requisite xml and rdf
header sections:
<?xml version="1.0"?>
<!DOCTYPE rdf:RDF [
<!ENTITY higgins
"http://www.eclipse.org/higgins/ontologies/2006/higgins">
<!ENTITY owl "http://www.w3.org/2002/07/owl# <http://www.w3.org/2002/07/owl>
">
<!ENTITY rdf "http://www.w3.org/1999/02/22- rdf-
syntax- ns#
<http://www.w3.org/1999/02/22- rdf-
syntax- ns> ">
<!ENTITY rdfs "http://www.w3.org/2000/01/rdf- schema#
<http://www.w3.org/2000/01/rdf- schema>
">
<!ENTITY xsd "http://www.w3.org/2001/XMLSchema#
<http://www.w3.org/2001/XMLSchema>
">
]>
<rdf:RDF xml:base="http://www.example.com/firstname"
xmlns:higgins="http://www.eclipse.org/higgins/ontologies/2006/higgins#
<http://www.eclipse.org/higgins/ontologies/2006/higgins>
"
xmlns:owl="&owl;"
xmlns:rdf="&rdf;" xmlns:rdfs="&rdfs"
xmlns:xsd="&xsd;">
## Here we are saying that we're defining a new
ontology that will describe
"first name" and that in turn imports
higgins.owl:
<owl:Ontology rdf:about="">
<owl:imports
rdf:resource="&higgins;.owl"/>
</owl:Ontology>
## The following says that "Firstname"
is a kind of
higgins:StringSimpleAttribute (i.e. a non-
compound (aka "simple") attribute
whose value is a string) and that it has exactly
one "firstnameValue" and
that firstname- dd object (defined further below)
is the resource that
provides the metadata to help display this
attribute and lastly that it is
the equivalent to "givenName" over at http://rivalattributes.com:
<owl:Class rdf:ID="Firstname">
<rdfs:subClassOf rdf:resource=
"&higgins;#StringSimpleAttribute"/>
<rdfs:subClassOf>
<owl:Restriction>
<owl:cardinality>1</owl:cardinality>
<owl:onProperty
rdf:resource="#firstnameValue"/>
</owl:Restriction>
</rdfs:subClassOf>
<higgins:displayData
rdf:resource="#firstname- dd"/>
<owl:equivalentClass rdf:resource=
"http://rivalattributes.com#givenName"/>
</owl:Class>
## The following says that the type of the value
of a Firstname attribute is
an XML Schema string with the restriction [a- zA-
Z]+ (copied from your
email):
<owl:DatatypeProperty rdf:ID="firstnameValue">
<rdfs:domain
rdf:resource="#Firstname"/>
<rdfs:range>
<rdfs:Datatype>
<higgins:base
rdf:resource="&xsd;string"/>
<higgins:pattern
rdf:datatype="&xsd;string">[a-
zA- Z]+</higgins:pattern>
</rdfs:Datatype>
</rdfs:range>
</owl:DatatypeProperty>
## The following says that the English language
display label is "First
Name". Other kinds of metadata are also
allowed here (but not shown)
including an image to associate with this
attribute and a
human- consumable/displayable description of the attribute:
<higgins:DisplayData
rdf:about="#firstname- dd">
<higgins:label xml:lang="en">First
Name</higgins:label>
</higgins:DisplayData>
## End of file
</rdf:RDF>
- Paul
PS: Just for fun, here's some more stuff that
shows how the above Firsname
attribute could be used in the definition of a
Person.
## A Person is a kind of Digital Subject...
<owl:Class rdf:ID="Person">
<rdfs:subClassOf
rdf:resource="&higgins;#DigitalSubject"/>
</owl:Class>
##...and that this Person has a firstname
predicate that connects it to an
instance of the FirstName class defined in the
email above. (The
sub- property element is optional, but nice to
have):
<owl:ObjectProperty
rdf:ID="firstname">
<rdfs:subPropertyOf
rdf:resource="&higgins;#attribute"/>
<rdfs:domain
rdf:resource="#Person"/>
<rdfs:range
rdf:resource="&higgins;#Firstname"/>
</owl:ObjectProperty>
_______________________________________________
higgins- dev mailing list
higgins- dev@xxxxxxxxxxx
https://dev.eclipse.org/mailman/listinfo/higgins- dev
_______________________________________________
higgins-dev mailing list
higgins-dev@xxxxxxxxxxx
https://dev.eclipse.org/mailman/listinfo/higgins-dev
