Re: Alignment of SSN and SOSA (completion of ACTION-246 Prepare a graphical representation of the alignment)

Hi,

I have added graphical representations of the proposed mappings (to serve
as basis for discussion for the group on issue-139)
This also corresponds to my ACTION-246: Prepare a graphical representation
of the alignment

The diagrams are drawn using a tool created by David Ratcliffe: Owl2Diagram
https://github.com/dratc/owl2diagram and are part of the ontology
documentation environment we use at ABS.

I have uploaded the images to this wiki page:
https://www.w3.org/2015/spatial/wiki/Mapping_Table

(uploading these images here is much faster/convenient for me than through
GitHub)
For convenience, I have added here the direct links to the corresponding
SVG files.

There are two sets of diagrams:

Simon's alignment (subClassOf)
https://github.com/w3c/sdw/blob/simon-ssn/ssn/rdf/ssn-sosa.ttl

... is a bit harder to read because the related stuff is connected by
arrows

[1] SOSA:
https://www.w3.org/2015/spatial/wiki/images/4/4f/Ontology-complete-sosa.svg

[2] SSN:
https://www.w3.org/2015/spatial/wiki/images/3/36/Ontology-complete-ssn.svg



... and Laurent's alignment (also based on the proposed mappings but opting
to implement them as owl:equivalentClass/owl:equivalentProperty)
--- is easier to read (and more compact) because the owl2diagram tool
creates a single box for equivalence classes.
(the mappings I have used are here:
https://www.w3.org/2015/spatial/wiki/images/d/db/Mappings.ttl


[3] SSN:
https://www.w3.org/2015/spatial/wiki/images/3/31/Ontology-complete-ssn-equiv.svg

[4] SOSA:
https://www.w3.org/2015/spatial/wiki/images/e/ec/Ontology-complete-sosa-equiv.svg

[5]  SSNX:
https://www.w3.org/2015/spatial/wiki/images/7/76/Ontology-complete-ssnx-equiv.svg

[6] OM Lite:
https://www.w3.org/2015/spatial/wiki/images/d/dc/Ontology-complete-oml-equiv.svg

 (using the seegrid.csiro .rdf version)
[7] Sampling Features Lite:
https://www.w3.org/2015/spatial/wiki/images/6/6c/Ontology-complete-samfl-equiv.svg

 (using the seegrid.csiro .rdf version)
and
[8] DUL:
https://www.w3.org/2015/spatial/wiki/images/9/9f/Ontology-complete-dul.svg


Note that I have done a number of changes (tweaks) to the source ontologies
to get images which are representative of the modelling intent.
In particular, I need the input to be OWL2 DL-compliant input because the
diagramming tool uses the Hermit reasoner, so I'm using repairing
strategies similar to the ones presented in the paper by N. Matentzoglu and
B. Parsia OWL Full/DL gap in the field
http://ceur-ws.org/Vol-1265/owled2014_submission_9.pdf
Plus there are other changes for prefixes (e.g. ssn moved to ssnx to
differentiate the old SSN to the new one).

Also, please note that this is not an endorsement of these alignments. My
position is similar to what was explained in the recent emails by Raul and
Raphael.
But it is easier to explain now that we have this. We should strive to:
- have a smaller number of boxes in the diagrams and also smaller boxes
(maximum two alternatives, not 3 or even 4 (Observation) as is currently
the case
- ensure maximum compatibility with the existing SSN
- ensure minimum overlap between the ontologies.
- and finally be capable to explain the added value associated to all the
changes we make (and motivate implementers ...).

-Laurent

PS: Phil, the other reason I have not tried via GitHub yet is because I
think I'm not yet declared as a contributor there (I was not around when
this was set up).



From: <Simon.Cox@csiro.au>
To: <public-sdw-wg@w3.org>,
Date: 24/01/2017 04:30 PM
Subject: Alignment of SSN and SOSA



I just created a little file containing an initial proposed mapping from
SSN to SOSA and loaded it to a branch – see

https://github.com/w3c/sdw/blob/simon-ssn/ssn/rdf/ssn-sosa.ttl


I just followed the alignment shown here:
https://www.w3.org/2015/spatial/wiki/Mapping_Table but expressed it more
formally so all the alignments are cast as

ssn:XXX rdfs:subClassOf sosa:YYY .
ssn:vvv rdfs:subPropertyOf sosa:uuu .

except for one, which is the other way round because the definition in SOSA
seemed to be more restricted than SSN.

sosa:hasValue rdfs:subPropertyOf ssn:hasValue ;
.

I think this matches the discussion today. I issued a pull-request.

In some cases it may be appropriate to replace rdfs:subClassOf with the
stronger owl:equivalentClass.

This relates to ISSUE-37, ISSUE-49, ISSUE-88, ISSUE-102, ISSUE-115,
ISSUE-139 .

Simon


From: Armin Haller [mailto:armin.haller@anu.edu.au]
Sent: Tuesday, 24 January, 2017 11:46
To: Kerry Taylor <kerry.taylor@anu.edu.au>; janowicz@ucsb.edu; Joshua
Lieberman <jlieberman@tumblingwalls.com>
Cc: Rob Atkinson <rob@metalinkage.com.au>; Raúl García Castro
<rgarcia@fi.upm.es>; public-sdw-wg@w3.org; Cox, Simon (L&W, Clayton)
<Simon.Cox@csiro.au>
Subject: Re: ACTION-251: (ISSUE-88) write up how an ssn:platform and a
sosa:platform are essentially the same, with an example (Spatial Data on
the Web Working Group)

I can see rapid convergence here in this thread. …