RE: ssn: issue-72 inverse properties in sosa-core

Hi Kerry -


?  not to a point of reaching  a  consensus on which  was preferred.

Not sure I understand that part of the assessment. There were multiple +1 for (A), and no votes against. There was only one +1 for option (B). (Danh voted 0 on both.) It looked like a reasonably strong consensus, though it was after the hour was up.

https://www.w3.org/2016/11/08-sdwssn-minutes

Simon

From: Kerry Taylor [mailto:kerry.taylor@anu.edu.au]
Sent: Wednesday, 9 November 2016 10:41 AM
To: SDW WG Public List <public-sdw-wg@w3.org>; Armin Haller <armin.haller@anu.edu.au>
Subject: ssn: issue-72 inverse properties in sosa-core

https://www.w3.org/2015/spatial/track/issues/72

The issue of inverse property declarations in sosa-core was again vehemently discussed in the meeting today. We got to a point where 2 alternative approaches  were universally considered acceptable if necessary, but not to a point of reaching  a  consensus on which  was preferred.

At the next ssn meeting this will be put to the vote.  The decision to be made is between  either A or B below.

The SOSA-core should often declare properties that are named, and intended to be used as inverses of other declared properties.
In every case where both a property and its intended  inverse is declared in sosa-core:

(A)    The pair is to be related by an owl:InverseOf declaration; or

(B)   The pair are not to be axiomatically  related but documentation is to be used to make the  inverse intention clear

Due to requests at the meeting the question of whether Classes on sosa-core  were to be declared as owl:Class or rdfs:Class was left open as it was suggested that that question was too overloaded with misunderstanding of the consequences, and focusing on  a particular question like inverseOf might set the scene. I guess the idea behind this is that if we think owl:InverseOf is ok to use in sosa-core --- a particular place where OWL reasoning can be used with a useful effect but without insisting that it *must* be used--  then we should also be comfortable with  using owl:Class in sosa-core.


--Kerry