Re: Non-dereferencable URIs (was Re: ACTION-255 - o&m alignment)

+1


On Wed, 8 Feb 2017 at 18:10 Krzysztof Janowicz <jano@geog.ucsb.edu> wrote:

> >I doubt this would be helpful but we could perhaps mint new ones that did
> dereference and assert >equivalence but that's only worth doing if it
> provides some functionality that's actually useful.
>
> I had the same (or similar) idea but it was already past 2 pm and we
> closed the meeting. What we could do is to have an ontology that has
> dereferencable URIs for all the classes and properties and for each of them
> would also contain rdfs:isDefinedBy statements that point to the O&M/ISO
> URIs. rdfs:isDefinedBy explicitly does not put any constraints on such
> resources, they do not even have to be Web-available (
> https://www.w3.org/TR/rdf-schema/#ch_isdefinedby). As a final step, we
> would establish our mapping axioms in the form of subclass and equivalence
> class relations from SOSA/SSN to these dereferenceable class and property
> URLs.
>
> Best,
> Jano
>
> On Tue, Feb 7, 2017 at 10:48 PM, Phil Archer <phila@w3.org> wrote:
>
> I'm sorry I wasn't able to speak about this on the call. By the time this
> topic came up I was being driven along the road and had insufficient
> bandwidth to un-mute myself although I could hear the conversation.
>
> Of course dereferencable URIs are preferred over non-dereferencable ones.
> However, this should not be seen as an absolute requirement or diktat. If
> there are good reasons to use non-dereferencable URIs - and it sounded to
> me as if Simon was making a very strong case for their use - and if the
> non-dereferencable state causes no harm, then go ahead.
>
> The question is, of course, can TC211 be encouraged to make things like
> http://def.isotc211.org/iso19156/2011/ dereferencable? If it helps, we'd
> happily host the relevant data to which those URIs could redirect (I'm sure
> OGC and others would too of course).
>
> I doubt this would be helpful but we could perhaps mint new ones that did
> dereference and assert equivalence but that's only worth doing if it
> provides some functionality that's actually useful.
>
> Phil
>
>
>
> On 07/02/2017 22:26, Simon.Cox@csiro.au wrote:
>
> I've dropped it into a Wiki page here (formatting not yet complete).
>
> https://www.w3.org/2015/spatial/wiki/Alignment_to_O%26M
>
> This should help with your namespace and prefix questions Laurent.
>
> From: Simon.Cox@csiro.au [mailto:Simon.Cox@csiro.au]
> Sent: Wednesday, 8 February, 2017 08:35
> To: public-sdw-wg@w3.org
> Subject: [ExternalEmail] RE: ACTION-255 - o&m alignment
>
> If you are unable to see the section in the HTML in the branch, here is
> the proposed text:
>
> -----------------------
>
> 8. Alignment to Observations and Measurements
>
> This section introduces the alignment of SOSA/SSN to OGC Observations and
> Measurements [OandM] (also known as ISO 19156:2011).
>
> O&M is specified as a UML model, following the patterns specified in ISO
> 19109 Geographic Information - Rules for Application Schema [ISO-19109].
> This means that the classes represent concepts from the application domain,
> so can be approximately equated with classes in an ontology.
>
> Two OWL implementations of O&M have been described:
>
> an explicit translation of the UML following the rules specified in
> [ISO-19150-2] - see [OM-Heavy]; and
> a handcrafted version in more idiomatic OWL [OM-Lite].
> The following sections provide two mappings or alignments between SOSA/SSN
> and O&M: the first with the official ISO/OGC UML conceptual model, and the
> second with the lightweight OWL implementation.
>
> 8.1 Alignment to Observations and Measurements UML model
>
> This section is non-normative.
>
> The explicit translation generates an RDF entity for every class, class
> attribute, and association-role from the original O&M UML model. It comes
> at a cost of a large set of dependencies on similar OWL translations of
> other UML models from the ISO 19100 series. Nevertheless, the URI for each
> RDF entity is a convenient identifier to elements of the UML model. These
> can be used to identify the elements of O&M in a formal RDF/OWL alignment.
>
> Rules for generating the URIs are provided in [ISO-19150-2], and appear in
> the official OWL implementation of ISO 19156 (O&M) maintained by the ISO/TC
> 211 Group on Ontology Management.
>
> The following namespace prefixes are used in the alignment to SOSA.
>
> Prefix   Namespace
> sosa:    http://www.w3.org/ns/sosa/
> sosa-om:          http://www.w3.org/ns/sosa-om#
> iso19156-gfi:
> http://def.isotc211.org/iso19156/2011/GeneralFeatureInstance#
> iso19156-om:  http://def.isotc211.org/iso19156/2011/Observation#
> iso19156-sf:     http://def.isotc211.org/iso19156/2011/SamplingFeature#
> iso19156-sfs:
> http://def.isotc211.org/iso19156/2011/SpatialSamplingFeature#
> iso19156-sp:    http://def.isotc211.org/iso19156/2011/Specimen#
> Utility classes
>
> Five utiity classes are defined locally to support the formalization of
> the alignment.
>
> 1. Three disjoint subclasses of sosa:Procedure:
>
> sosa-om:ActuationProcedure Actuation procedures or recipes
> sosa-om:ObservationProcedure         Observation procedures or recipes
> sosa-om:SamplingProcedure  Sampling, sample preparation or processing
> procedures or recipes
> 2. Two classes related to sampling, which complement SOSA classes related
> to actuation and observation:
>
> sosa-om:SamplingDevice        Sampling, sample preparation or processing
> devices, comparable to sosa:Actuator and sosa:Sensor
> sosa-om:SamplingEvent          Sampling, sample preparation or processing
> event or act, comparable to sosa:Actuation and sosa:Observation
> Class alignments
>
> The primary classes from [OandM] have direct equivalents in SOSA classes
> supplemented by the utility classes described above, as follows:
>
> iso19156-om:OM_Observation          equivalent class
> sosa:Observation
> iso19156-om:OM_Process      equivalent class          sosa:Sensor or
> sosa-om:ObservationProcedure
> iso19156_sf:SF_SamplingFeature       equivalent class          sosa:Sample
> iso19156-sf:SF_Process           equivalent class
> sosa-om:SamplingDevice or sosa-om:SamplingProcedure
> Additional alignments from SOSA/SSN classes to O&M classes are as follows.
>
> sosa:FeatureOfInterest           subclass of
>  iso19156_gfi:GFI_DomainFeature
> where iso19156_gfi:GFI_DomainFeature has the definition:
>
> The class GFI_DomainFeature represents 'real-world' features which are the
> ultimate subject of an observation campaign, i.e. the features from an
> application domain that are not artefacts of the observation process
> (sampling features).
> sosa:Actuator  subclass of       iso19156_gfi:GFI_Feature
> sosa:Platform  subclass of       iso19156_gfi:GFI_Feature
> where iso19156_gfi:GFI_Feature has the definition
>
> The class GFI_Feature represents the set of all classes which are feature
> types. In an implementation this abstract class shall be substituted by a
> concrete class representing a feature type from an application schema
> associated with a domain of discourse (ISO 19109, ISO 19101).
> Property alignments
>
> The following properties from [OandM] have direct equivalents in SOSA
> properties:
>
> iso19156-om:OM_Observation.featureOfInterest     equivalent property
> sosa:hasFeatureOfInterest
> iso19156-om:OM_Observation.observedProperty    equivalent property
> sosa:observedProperty
> iso19156-om:OM_Observation.phenomenonTime    equivalent property
> sosa:phenomenonTime
> iso19156-sf:SF_SamplingFeature.sampledFeature    equivalent property
> sosa:isSampleOf
> Additional alignments from O&M properties to SOSA are as follows.
>
> iso19156-om:OM_Observation.procedure    sub-property of
>  sosa:usedProcedure
> iso19156-sp:SF_Specimen.samplingMethod  sub-property of
>  sosa:usedProcedure
> These alignments are modeled as sub-properties because sosa:usedProcedure
> applies to actuation, observation or sampling activities.
>
> iso19156-om:OM_Observation.result            sub-property of
>  sosa:hasResult
> iso19156-om:OM_Observation.resultTime    sub-property of
>  sosa:resultTime
> iso19156-sp:SF_Specimen.samplingTime       sub-property of
>  sosa:resultTime
> iso19156-sp:PreparationStep.time     sub-property of
>  sosa:resultTime
> These alignments are modeled as sub-properties because sosa:hasResult and
> sosa:resultTime applies to actuation, observation or sampling activities.
>
> iso19156-sfs:SF_SpatialSamplingFeature.hostedProcedure   sub-property of
>          sosa:hosts
> These alignments are modeled as a sub-property because the domain of
> iso19156-sfs:SF_SpatialSamplingFeature.hostedProcedure is a spatial
> sampling feature, such as a station, rather than a more general platform.
>
> An RDF file containing a graph corresponding to this alignment is
> available.
>
> 8.2 Alignment to om-lite implementation of Observations and Measurements
>
> This section is non-normative.
>
> An idiomatic OWL implementation of O&M (including Sampling Features) is
> described in [OM-Lite].
>
> The following namespace prefixes are used in the alignment to SOSA.
>
> Prefix   Namespace
> sosa:    http://www.w3.org/ns/sosa/
> sosa-oml:         http://www.w3.org/ns/sosa-oml#
> oml:     http://def.seegrid.csiro.au/ontology/om/om-lite#
> samfl:  http://def.seegrid.csiro.au/ontology/om/sam-lite#
> Utility classes
>
> Five utiity classes are defined locally to support the formalization of
> the alignment.
>
> 1. Three disjoint subclasses of sosa:Procedure:
>
> sosa-oml:ActuationProcedure Actuation procedures or recipes
> sosa-oml:ObservationProcedure        Observation procedures or recipes
> sosa-oml:SamplingProcedure Sampling, sample preparation or processing
> procedures or recipes
> 2. Two classes related to sampling, which complement SOSA classes related
> to actuation and observation:
>
> sosa-oml:SamplingDevice       Sampling, sample preparation or processing
> devices, comparable to sosa:Actuator and sosa:Sensor
> sosa-oml:SamplingEvent         Sampling, sample preparation or processing
> event or act, comparable to sosa:Actuation and sosa:Observation
> Class alignments
>
> The primary classes from [OM-Lite] have direct equivalents in SOSA classes
> supplemented by the utility classes described above, as follows:
>
> oml:Observation         equivalent class          sosa:Observation
> oml:Process     equivalent class          sosa:Sensor or
> sosa-om:ObservationProcedure
> samfl:SamplingFeature           equivalent class          sosa:Sample
> samfl:Process  equivalent class          sosa-om:SamplingDevice or
> sosa-om:SamplingProcedure
> Property alignments
>
> The following properties from [OM-Lite] have direct equivalents in SOSA
> properties:
>
> oml:featureOfInterest equivalent property    sosa:hasFeatureOfInterest
> oml:observedProperty            equivalent property
> sosa:observedProperty
> oml:phenomenonTime           equivalent property    sosa:phenomenonTime
> samfl:sampledFeature            equivalent property    sosa:isSampleOf
> Additional alignments from [OM-Lite] properties to SOSA are as follows.
>
> oml:procedure            sub-property of           sosa:usedProcedure
> samfl:samplingMethod           sub-property of           sosa:usedProcedure
> These alignments are modeled as sub-properties because sosa:usedProcedure
> applies to actuation, observation or sampling activities.
>
> oml:result        sub-property of           sosa:hasResult
> oml:resultTime           sub-property of           sosa:resultTime
> samfl:samplingTime   sub-property of           sosa:resultTime
> These alignments are modeled as sub-properties because sosa:hasResult and
> sosa:resultTime applies to actuation, observation or sampling activities.
>
> samfl:hostedProcedure           sub-property of           sosa:hosts
> These alignments are modeled as a sub-property because the domain of
> iso19156-sfs:SF_SpatialSamplingFeature.hostedProcedure is a spatial
> sampling feature, such as a station, rather than a more general platform.
>
> An RDF file containing a graph corresponding to this alignment is
> available.
>
>
> From: Cox, Simon (L&W, Clayton)
> Sent: Tuesday, 7 February, 2017 15:58
> To: 'Simon.Cox@csiro.au' <Simon.Cox@csiro.au<mailto:Simon.Cox@csiro.au>>;
> public-sdw-wg@w3.org<mailto:public-sdw-wg@w3.org>
> Subject: RE: ACTION-255 - o&m alignment
>
> Note that this mapping is only from SOSA to O&M (and om-lite).
>
> I also intend to look at the mapping from 'SSN' (i.e. the vertical
> axiomatization/extension) of SOSA for observations and sensing through to
> O&M. Much of the intended alignment was captured in annotations in the
> original ontology, but the discussions in the last week suggest that some
> local-names might change else SOSA classes and properties used in place of
> old SSN equivalents. When this has settled down a comprehensive mapping
> should be formulated.
>
> Probably the key question arising from the work mentioned below is whether
> the proposal to use the URIs that are specified in the translation of the
> UML model to OWL, following the ISO 19150-2 rules, is acceptable.
>
> Pro:
>
> -          It allows us to express the alignment formally within the idiom
> we are working in (OWL)
> Cons:
>
> -          The OWL version of O&M is not in itself published as a
> 'standard' and the URIs are not directly resolvable (yet, anyway)
>
> o   However, as pointed out in the document, the URIs are persistent
> identifiers in the view of ISO, and the fact that ISO's process does not
> require a separate document for the OWL implementation should be OK for us
>
> -          UML and OWL are so different in their assumptions that it is a
> fallacy to use the OWL implementation as representing the UML model
>
> I'm sure there are other arguments. My feeling is that the proposed
> approach balances formality and pragmatism OK.
>
> Simon
>
> From: Simon.Cox@csiro.au<mailto:Simon.Cox@csiro.au> [mailto:
> Simon.Cox@csiro.au]
> Sent: Saturday, 28 January, 2017 21:33
> To: public-sdw-wg@w3.org<mailto:public-sdw-wg@w3.org>
> Subject: [ExternalEmail] ACTION-255 - o&m alignment
>
> In response to my ACTION-255 from last week's meeting, I have generated
> two RDF files containing formal alignments between SOSA and O&M.
>
> -          sosa-om-mapping.ttl relates to the O&M UML model, and uses the
> Official ISO URIs from the OWL implementation prepared by the ISO/TC 211
> Group on Ontology Management following the rules from ISO 19150-2
>
> -          sosa-oml-mapping.ttl relates to the om-lite and sam-lite OWL
> implementation recently published in Semantic Web Journal
>
> I have also prepared text for the SSN document describing these mappings -
> for chapter 8 in the spec.
> So far the mappings only concern SOSA.
>
> I've pushed all this into a branch in GitHub
> https://github.com/w3c/sdw/tree/simon-ssn-O%26M-alignments/ssn
> and issued a pull-request https://github.com/w3c/sdw/pull/516
>
> Simon
>
> Simon J D Cox
> Research Scientist
> Environmental Informatics
> CSIRO Land and Water<http://www.csiro.au/Research/LWF>
>
> E simon.cox@csiro.au<mailto:simon.cox@csiro.au> T +61 3 9545 2365 M +61
> 403 302 672
>    Mail: Private Bag 10, Clayton South, Vic 3169
>    Visit: Central Reception, Research Way, Clayton, Vic 3168
>    Deliver: Gate 3, Normanby Road, Clayton, Vic 3168
> people.csiro.au/Simon-Cox<http://people.csiro.au/Simon-Cox>
> orcid.org/0000-0002-3884-3420<http://orcid.org/0000-0002-3884-3420>
> researchgate.net/profile/Simon_Cox3<
> https://www.researchgate.net/profile/Simon_Cox3>
> github.com/dr-shorthair<https://github.com/dr-shorthair>
>
> PLEASE NOTE
> The information contained in this email may be confidential or privileged.
> Any unauthorised use or disclosure is prohibited. If you have received this
> email in error, please delete it immediately and notify the sender by
> return email. Thank you. To the extent permitted by law, CSIRO does not
> represent, warrant and/or guarantee that the integrity of this
> communication has been maintained or that the communication is free of
> errors, virus, interception or interference.
>
> Please consider the environment before printing this email.
>
>
>
>
> --
>
>
> Phil Archer
> Data Strategist, W3C
> http://www.w3.org/
>
> http://philarcher.org
> +44 (0)7887 767755
> @philarcher1
>
>
>