[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: RE: [tag] My Conclusion re An RDF/S/OWL-DL representation
On a topic related to TA precedence, but unrelated to RDF:
I happened to do further work on WS-I testing technology, and could
implement the "prerequisite" relationship efficiently enough thanks to
XSLT2.0.
The WS-I context is rather narrow for our purpose: Tas are supposed to
target only XML artifacts and be expressed (unlike in past versions)
only using XPath2.0 expressions, meaning the selection of the target,
and the predicate, are XPath expr. So in this very particular case, the
TA itself is executable.
Incidentally, this is a model very close to Schematron (which is an ISO
standard), so it is still representative of a significant application
space. However Schematron does a poor job at handling relationships
across Tas and has no support for prerequisites.
For WS-I, a simple XSLT2.0 script did the job perfectly - executing any
set of WS-I TAs according to the partial order defined by the
prerequisite relationship, over a given test input (an XML container
aggregating WSDL artifacts and message traces). No doubt that all sorts
of TA groupings (including embedded groups) canbe supported too.
A demo at the next F2F could be of interest.
My plan is, once we have defined a mark-up for Tas, to make a TA
execution engine in XSLT that would directly execute them in the case
these Tas use Xpath for specifying targets and predicates.
So after the guideline doc is completed, some exciting work ahead...
Jacques
-----Original Message-----
From: stephen.green@systml.co.uk [mailto:stephen.green@systml.co.uk]
Sent: Wednesday, January 30, 2008 11:18 AM
To: tag@lists.oasis-open.org
Subject: [tag] My Conclusion re An RDF/S/OWL-DL representation
The overlap of requirements addressed by RDF/S/OWL and the requirements
of a TAG markup are very interesting and make RDF/S/OWL attractive for
marking up a Test Assertion. However I think there are a few very
significant mismatches (as far as I understand it):
* OWL and to some extent RDF/S have been designed for an 'Open System'
where any set of assertions can be mixed with any other
* there is no weighting of one assertion over another, each have equal
weight
* the set of assertions stated with RDF/S/OWL cannot be closed to
further
assertions or given precedence over other assertions
* one assertion in OWL cannot replace a previous one, only supplement it
I feel these and a few other features of RDF/S/OWL make it on the whole
a poor match for some of the important aspects of test assertions and so
I wouldn't particularly favor its use for TAs. Mostly this is due to the
need to designate a set of assertions as being in some way complete and
authoritative and part of a particular group. Of course this is a matter
of degree in that further assertions might well be added, but not as
freely, I feel, as OWL would allow.
Further developments in the Semantic Web do seem to be addressing such
less-open system requirements but it remains to be seen whether this
will lead to something useful enough for test assertions, I think.
--
Stephen D. Green
Partner
SystML, http://www.systml.co.uk
Tel: +44 (0) 117 9541606
http://www.biblegateway.com/passage/?search=matthew+22:37 .. and voice
Quoting stephen.green@systml.co.uk:
> A roundtrip works quite well, to get an idea of how others might use
> RDF/RDFS/OWL to represent the same Test Assertion model/example. An
> export from the semantic wiki Halo project (which imports well back
> into the wiki) looks like this for the exact same example:
>
> <?xml version="1.0" encoding="UTF-8"?> <!DOCTYPE rdf:RDF[
> <!ENTITY rdf 'http://www.w3.org/1999/02/22-rdf-syntax-ns#'>
> <!ENTITY rdfs 'http://www.w3.org/2000/01/rdf-schema#'>
> <!ENTITY owl 'http://www.w3.org/2002/07/owl#'>
> <!ENTITY smw 'http://smw.ontoware.org/2005/smw#'>
> <!ENTITY smwdt 'http://smw.ontoware.org/2005/smw-datatype#'>
> <!ENTITY wiki
'http://localhost/mediawiki/index.php/Special:URIResolver/'>
> <!ENTITY wikiurl 'http://localhost/mediawiki/index.php/'>
> <!ENTITY property
> 'http://localhost/mediawiki/index.php/Special:URIResolver/Property-3A/
> '>
> ]>
>
> <rdf:RDF
> xmlns:rdf="&rdf;"
> xmlns:rdfs="&rdfs;"
> xmlns:owl ="&owl;"
> xmlns:smw="&smw;"
> xmlns:wiki="&wiki;"
> xmlns:wikiurl="&wikiurl;"
> xmlns:property="&property;">
> <!-- Ontology header -->
> <owl:Ontology rdf:about="">
> <smw:exportDate
>
rdf:datatype="http://www.w3.org/2001/XMLSchema#dateTime";>2008-01-28T10:0
0:16+00:00</smw:exportDate>
> </owl:Ontology>
> <!-- exported page data -->
> <smw:Thing rdf:about="&wiki;Widget_precondition1">
> <rdfs:label>Widget precondition1</rdfs:label>
>
> <smw:hasArticle
rdf:resource="&wikiurl;Widget_precondition1"/>
> <rdfs:isDefinedBy
> rdf:resource="&wikiurl;Special:ExportRDF/Widget_precondition1"/>
> <rdf:type rdf:resource="&wiki;Category-3AWidget"/>
> <property:Property
rdf:resource="&wiki;WidgetSize_Medium"/>
> </smw:Thing>
> <smw:Thing rdf:about="&wiki;Widget_ta001">
> <rdfs:label>Widget ta001</rdfs:label>
> <smw:hasArticle rdf:resource="&wikiurl;Widget_ta001"/>
>
> <rdfs:isDefinedBy
rdf:resource="&wikiurl;Special:ExportRDF/Widget_ta001"/>
> <rdf:type rdf:resource="&wiki;Category-3AWidget"/>
> <property:Prose
> rdf:datatype="http://www.w3.org/2001/XMLSchema#string";>has a red
> button on top</property:Prose>
> <property:SpecRef
>
rdf:datatype="http://www.w3.org/2001/XMLSchema#string";>101:1a</property:
SpecRef>
> <property:Precondition
rdf:resource="&wiki;Widget_precondition1"/>
> </smw:Thing>
> <!-- auxilliary definitions -->
> <owl:ObjectProperty rdf:about="&property;Precondition">
>
> <rdfs:label>Precondition</rdfs:label>
> <smw:hasArticle
rdf:resource="&wikiurl;Property:Precondition"/>
> <rdfs:isDefinedBy
> rdf:resource="&wikiurl;Special:ExportRDF/Property:Precondition"/>
> </owl:ObjectProperty>
> <owl:DatatypeProperty rdf:about="&property;SpecRef">
> <rdfs:label>SpecRef</rdfs:label>
> <smw:hasArticle
rdf:resource="&wikiurl;Property:SpecRef"/>
> <rdfs:isDefinedBy
> rdf:resource="&wikiurl;Special:ExportRDF/Property:SpecRef"/>
>
> </owl:DatatypeProperty>
> <owl:DatatypeProperty rdf:about="&property;Prose">
> <rdfs:label>Prose</rdfs:label>
> <smw:hasArticle rdf:resource="&wikiurl;Property:Prose"/>
> <rdfs:isDefinedBy
> rdf:resource="&wikiurl;Special:ExportRDF/Property:Prose"/>
> </owl:DatatypeProperty>
> <smw:Thing rdf:about="&wiki;WidgetSize_Medium">
> <rdfs:label>WidgetSize Medium</rdfs:label>
>
> <smw:hasArticle
rdf:resource="&wikiurl;WidgetSize_Medium"/>
> <rdfs:isDefinedBy
> rdf:resource="&wikiurl;Special:ExportRDF/WidgetSize_Medium"/>
> </smw:Thing>
> <owl:ObjectProperty rdf:about="&property;Property">
> <rdfs:label>Property</rdfs:label>
> <smw:hasArticle
rdf:resource="&wikiurl;Property:Property"/>
> <rdfs:isDefinedBy
> rdf:resource="&wikiurl;Special:ExportRDF/Property:Property"/>
> </owl:ObjectProperty>
>
> <owl:Class rdf:about="&wiki;Category-3AWidget">
> <rdfs:label>Widget</rdfs:label>
> <smw:hasArticle
rdf:resource="&wikiurl;Category:Widget"/>
> <rdfs:isDefinedBy
> rdf:resource="&wikiurl;Special:ExportRDF/Category:Widget"/>
> </owl:Class>
> <!-- reference to the Semantic MediaWiki schema -->
> <owl:AnnotationProperty rdf:about="&smw;hasArticle">
> <rdfs:isDefinedBy
rdf:resource="http://smw.ontoware.org/2005/smw"/>
>
> </owl:AnnotationProperty>
> <owl:AnnotationProperty rdf:about="&smw;exportDate">
> <rdfs:isDefinedBy
rdf:resource="http://smw.ontoware.org/2005/smw"/>
> </owl:AnnotationProperty>
> <owl:Class rdf:about="&smw;Thing">
> <rdfs:isDefinedBy
rdf:resource="http://smw.ontoware.org/2005/smw"/>
> </owl:Class>
> <!-- Created with Semantic MediaWiki,
http://ontoworld.org/wiki/SMW
> --> </rdf:RDF>
>
> --
> Stephen D. Green
>
> Partner
> SystML, http://www.systml.co.uk
> Tel: +44 (0) 117 9541606
>
> http://www.biblegateway.com/passage/?search=matthew+22:37 .. and voice
>
>
>
> Quoting stephen.green@systml.co.uk:
>
>> Attached is a screen dump of how this representation looks in the
>> semantic wiki. It revealed a few modeling issues...
>>
>> [Incidentally there was a missing spec ref - it should look like
this:
>>
>> <Widget rdf:ID="Widget_precondition1">
>> <property>
>> <WidgetSize rdf:ID="WidgetSize_Medium"/>
>> </property>
>> </Widget>
>> <Widget rdf:ID="Widget_ta001">
>> <specRef
>>
rdf:datatype="http://www.w3.org/2001/XMLSchema#string";>101:1a</specRef>
>> <prose
rdf:datatype="http://www.w3.org/2001/XMLSchema#string";>has a
>> red button on top</prose>
>> <precondition rdf:resource="#Widget_precondition1"/>
>> </Widget>
>> ]
>>
>> There is an issue to note, I think, with the use of a 'precondition'
>> - it is a bit of a modeling problem in that the precondition actually
>> functions as a restricted subclass of the test assertion or its
predicate.
>>
>> * If the precondition is itself a predicate it needs everything
>> a test assertion needs except the spec ref. I.e. maybe it needs
>> its own ID so one precondition can be used in multiple predicates.
>>
>> * This then leads to some issues:
>>
>> * Why not then just use an assertion?
>>
>> * What then is the difference between precondition and prerequisite?
>>
>> * Doesn't this logic mean that a test assertion actually amounts
>> an extension of a superclass which is an assertion (extended by
>> adding spec ref)?
>> * Then a prerequisite could be the equivalent of a precondition
>> in that both reference both assertions and test assertions (or
>> instantiate, rather than reference the assertion and test
>> assertion classes).
>>
>> * So having a new superclass of test assertion called 'assertion'
>> could answer three problems:
>> 1. no need for both prerequisites and preconditions
>> 2. distinction between
>> a. test assertions and
>> b. assertions/predicates which aren't test assertions as
such
>> 3. a different target can be associated with the
assertion/predicate
>> of the precondition/prerequisite - a different target to the
>> TA's
>>
>> Another issue arose while trying to implement the model:
>>
>> Does a TA have a target or does a target have a TA?
>>
>> The latter proves to be most useful in my example (to allow grouping
>> of TAs under the category of a shared target or target class).
>>
>> Best regards
>> - Steve
>>
>>> I tried the OWL-DL representation with WikiMedia's SemanticWiki /
>>> Halo Project OWL extensions and found that it needed tightening up a
>>> lot more (eliminating the redundant elements).
>>>
>>> This resulted in the representation below which seems quite
>>> satisfactory as a mere demo of how a TA might be represented for the
semantic web:
>>>
>>> <?xml version="1.0"?>
>>> <rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#";
>>> xmlns:owl="http://www.w3.org/2002/07/owl#"; xmlns="http://localhost#";
>>> xmlns:rdfs="http://www.w3.org/2000/01/rdf-schema#";
>>> xml:base="http://localhost";>
>>> <owl:Ontology rdf:about=""/>
>>> <owl:Class rdf:ID="Property"/>
>>> <owl:Class rdf:ID="WidgetSize">
>>> <rdfs:subClassOf>
>>> <owl:Class rdf:ID="Size"/>
>>> </rdfs:subClassOf>
>>> </owl:Class>
>>> <owl:Class rdf:ID="Target"/>
>>> <owl:Class rdf:about="#Size">
>>> <rdfs:subClassOf rdf:resource="#Property"/>
>>> </owl:Class>
>>> <owl:Class rdf:ID="Widget">
>>> <rdfs:subClassOf rdf:resource="#Target"/>
>>> </owl:Class>
>>> <owl:ObjectProperty rdf:ID="precondition">
>>> <rdfs:domain rdf:resource="#Target"/>
>>> </owl:ObjectProperty>
>>> <owl:DatatypeProperty rdf:ID="prerequisite">
>>> <rdfs:domain rdf:resource="#Target"/>
>>> <rdfs:range
rdf:resource="http://www.w3.org/2001/XMLSchema#string"/>
>>> </owl:DatatypeProperty>
>>> <owl:DatatypeProperty rdf:ID="specRef">
>>> <rdfs:domain rdf:resource="#Target"/>
>>> <rdfs:range
rdf:resource="http://www.w3.org/2001/XMLSchema#string"/>
>>> </owl:DatatypeProperty>
>>> <owl:FunctionalProperty rdf:ID="property">
>>> <rdfs:domain rdf:resource="#Target"/>
>>> <rdfs:range rdf:resource="#Property"/>
>>> <rdf:type
rdf:resource="http://www.w3.org/2002/07/owl#ObjectProperty"/>
>>> </owl:FunctionalProperty>
>>> <owl:FunctionalProperty rdf:ID="value">
>>> <rdfs:domain rdf:resource="#Property"/>
>>> <rdfs:range
rdf:resource="http://www.w3.org/2001/XMLSchema#string"/>
>>> <rdf:type
rdf:resource="http://www.w3.org/2002/07/owl#DatatypeProperty"/>
>>> </owl:FunctionalProperty>
>>> <owl:FunctionalProperty rdf:ID="prose">
>>> <rdfs:range
rdf:resource="http://www.w3.org/2001/XMLSchema#string"/>
>>> <rdfs:domain rdf:resource="#Target"/>
>>> <rdf:type
rdf:resource="http://www.w3.org/2002/07/owl#DatatypeProperty"/>
>>> </owl:FunctionalProperty>
>>> <Widget rdf:ID="Widget_precondition1">
>>> <property>
>>> <WidgetSize rdf:ID="WidgetSize_Medium"/>
>>> </property>
>>> </Widget>
>>> <owl:AllDifferent/>
>>> <Widget rdf:ID="Widget_ta001">
>>> <specRef
rdf:datatype="http://www.w3.org/2001/XMLSchema#string"/>
>>> <prose
rdf:datatype="http://www.w3.org/2001/XMLSchema#string";>has
>>> a red button on top</prose>
>>> <precondition rdf:resource="#Widget_precondition1"/>
>>> </Widget>
>>> </rdf:RDF>
>>> <!-- Created with Protege (with OWL Plugin 3.3.1, Build 430)
>>> http://protege.stanford.edu -->
>>>
>>>
>>> --
>>> Stephen D. Green
>>>
>>> Partner
>>> SystML, http://www.systml.co.uk
>>> Tel: +44 (0) 117 9541606
>>>
>>> http://www.biblegateway.com/passage/?search=matthew+22:37 .. and
>>> voice
>>>
>>>
>>>
>>> Quoting stephen.green@systml.co.uk:
>>>
>>>> Second attempt after first seems to have got stuck (in filters?)
>>>> This time with attachment as zip
>>>>
>>>> Contains two OWL files, one with a TA ontology - classes and
>>>> properties, the other with an example individual TA in OWL-DL using
the ontology.
>>>>
>>>> --
>>>> Stephen D. Green
>>>>
>>>> Partner
>>>> SystML, http://www.systml.co.uk
>>>> Tel: +44 (0) 117 9541606
>>>>
>>>> http://www.biblegateway.com/passage/?search=matthew+22:37 .. and
>>>> voice
>>>
>>>
>
>
>
>
> ---------------------------------------------------------------------
> To unsubscribe from this mail list, you must leave the OASIS TC that
> generates this mail. You may a link to this group and all your TCs in
> OASIS
> at:
> https://www.oasis-open.org/apps/org/workgroup/portal/my_workgroups.php
---------------------------------------------------------------------
To unsubscribe from this mail list, you must leave the OASIS TC that
generates this mail. You may a link to this group and all your TCs in
OASIS
at:
https://www.oasis-open.org/apps/org/workgroup/portal/my_workgroups.php
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]