Re: Comment on the SOSA ontology

Dear Catherine,

Thank you for your comments.

In a nutshell, according to the original SSN, a feature of interest would
be "plot X", and a property would not be generic but would be specific to
"plot X", like: "temperature of soil at 3 meter depth in plot X".

see also section 5.3.1.3.4 in
https://www.w3.org/2005/Incubator/ssn/XGR-ssn-20110628/#Observed_Properties_2


*Observed Properties:* ObservedProperty is defined as a subclass of
DUL:Quality. Types of properties, such as temperature or pressure should be
added as subclasses of ObservedProperty instead of individuals. A new
relation called SSO:isPropertyOf is defined as a subrelation of
DUL:isQualityOf to relate a property to a feature of interest.


Hence generic Feature of Interest and Properties are modeled as classes:
"the class of plot", or "the class of temperatures of soil at 3 meter
depth". A repository in the agriculture domain would define *classes* of
feature of interest and *classes* of properties that may apply to them, and
you would instantiate these classes in datasets.


On the other hand, in existing SSN datasets, instances of oldssn:Property
are often generic and can apply to more than one feature of interest.
Although it contradicts the original specification, this way of modeling
the domain seems to have become the post-facto standard.


This has been extensively discussed in the group lately, see also:
-
https://www.w3.org/2015/spatial/wiki/Link_between_FeatureOfInterest_and_xxxProperty#What_is_an_instances_of_ssn:Property_.3F
- https://lists.w3.org/Archives/Public/public-sdw-wg/2017May/0129.html
- https://lists.w3.org/Archives/Public/public-sdw-wg/2017Apr/0336.html
- https://lists.w3.org/Archives/Public/public-sdw-wg/2017May/0098.html
<https://lists.w3.org/Archives/Public/public-sdw-wg/2017Apr/0336.html>


We will hopefully be able to decide on this late issue during our meeting
tomorrow. Anyways, soon after the decision is made we will enhance the
examples to better illustrate how features of interest, properties, and
samples, can play together.

Kind regards,
Maxime Lefrançois



Le lun. 15 mai 2017 à 10:18, Catherine Roussey <catherine.roussey@irstea.fr>
a écrit :

> Thanks for the information
>
> I have found how to define the unit thanks to your QUDT example.
>
> I still find strange the way you use feature of interets and property.
>
> It seems as far as I understand that the property belong to the instance
> of the feature of interest and are not generic.
>
> in the appartement example electrical consumption may be apply to
> different type of feature of interest so I found the URI of the property
> quite strange.
>
> Why in the URI of property there is an indication of the instance
> appartment/134/ electricalconsumption or tree/124#height
>
> I was expecting that at the end there will exist some repositoryof
> instance per domain that list all the possible feature of interest and
> property. The sample will help to identify exactly how the property of the
> feature of interest is measured.
>
> for example in the agriculture domain, I would expect that we define a
> repository  per agriculture domain :
>
> soil temperature and humidity at 3 meter or 6 meter depth (whatever the
> plot is): feature of interest (soil at 3 meter depth) property (temperature)
>
> air temperature and humidity at 2 meter  height (wherever we are):
> featureof interest (outdoor air  around 2 meter height) property
> (temperature)
>
> air temperature and humidity of a breeding building...(for any breeding
> building): feature of interest (indoor air) property (temperature)
>
> The repository is generic so the sample will help to identify where the
> sensor are located and what is measured exactly.
>
> But as far as I understand your example
>
> you will imagine that we will change the repository of instances to be
> more precise. A repository per specific farm.
>
> . The property is linked to the instance of feature of interest.
>
>  temperature and humidity at 3 meter or 6 meter depth of the  plot soil
> located at .... (how  can we access the measure with a query about soil
> temperature at 3 meter depth?)
>
> So at the moment I do not see how feature of interest, property and sample
> can play together.
>
> I miss something.
>
>
> <https://www.w3.org/2015/spatial/track/actions/349%29>
> Best Regards <https://www.w3.org/2015/spatial/track/actions/349%29>
>
> Catherine
> <https://www.w3.org/2015/spatial/track/actions/349%29>
>
>
> Le 15/05/2017 à 08:44, Armin Haller a écrit :
>
> Hi Catherine,
>
>
>
> Thanks for your additional comments. The examples on Github is a bit
> outdated and we are currently working on a new example. It is currently
> undergoing review: https://github.com/w3c/sdw/pull/891. The issue you
> raise regarding what is a FeatureOfInterest and what is the Property is
> currently being discussed in the group (see also
> https://www.w3.org/2015/spatial/track/actions/349). The hasProperty and
> isPropertyOf relations that relate a Property to a FeatureOfInterest are
> currently only defined in SSN, but we are considering their inclusion in
> the SOSA core.
>
>
>
> The hasValue property has actually been replaced by the hasSimpleResult
> property which can be used to attach a value to an Observation/Actuation.
> If units of measurement are attached, the use of the sosa:Result class is
> required. For units of measurement SOSA as well as SSN are relying on
> external ontologies such as the QUDT. However, we will include an example
> how to attach a unit of measure to a sosa:Result using an external ontology
> such as QUDT.
>
>
>
> Kind regards,
> Armin
>
>
>
> *From: *Catherine Roussey <catherine.roussey@irstea.fr>
> <catherine.roussey@irstea.fr>
> *Date: *Wednesday, 10 May 2017 at 8:25 pm
> *To: *"public-sdw-comments@w3.org" <public-sdw-comments@w3.org>
> <public-sdw-comments@w3.org> <public-sdw-comments@w3.org>
> *Subject: *Re: Comment on the SOSA ontology
> *Resent-From: *<public-sdw-comments@w3.org> <public-sdw-comments@w3.org>
> *Resent-Date: *Wednesday, 10 May 2017 at 8:26 pm
>
>
>
> Dear all
>
> I am some remark about the sosa example
>
> https://github.com/w3c/sdw/blob/gh-pages/ssn/rdf/documentation_examples/sosa-core_examples.ttl
>
>
> In the ssn ontology it was  clear what is a FeatureOfInterest (wind) and a
> Property of a FeatureOfInterest ( wind direction).
> This distinction is no so easy to understand at the first reading. That
> why the design patern SSO is so important.
> But I like very much this distinction. A temperature sensor may measure
> different think depending of its location (the soil, the outdoor air, the
> indoor air)
>
> About the sosa example of room temperature
> for me the FeatureOfInterest is a room (or buiding or indoor air),
> identified by one sample (a specific room 4830EH_UCSB)
> and the property is the temperature (room temperature ).
> In the example I think it is not so clear because I can not see where is
> defined  the property (temperature) .
>
> the result of observation has no unit.
> sosa-core:hasValue "23.8"^^xsd:double ;
> there exist different unit about temperature celsius or degree.
> Where can we find the unit of the temperature value?
>
>
> Best regards
> Catherine
>
> --
>
> Catherine ROUSSEY
>
> Irstea Clermont Ferrand
>
> Campus des Cézeaux
>
> 9 avenue Blaise Pascal
>
> CS 200 85
>
> 63178 Aubière
>
> tel: 33 (0)4 73 44 06 88 <04%2073%2044%2006%2088>
>
> "Imprefection si beauty, madness is genious and it's better to be absolutely ridiculous than aboslutely boring" Maryline Monroe
>
>
> --
> Catherine ROUSSEY
> Irstea Clermont Ferrand
> Campus des Cézeaux
> 9 avenue Blaise Pascal
> CS 200 85
> 63178 Aubière
> tel: 33 (0)4 73 44 06 88 <04%2073%2044%2006%2088>
> "Imprefection si beauty, madness is genious and it's better to be absolutely ridiculous than aboslutely boring" Maryline Monroe
>
>