Re: Comment on the SOSA ontology


Hi Nicolas,

Thank you for your interest in SOSA and the comments you make below. We have raised the lack of examples as an issue and we are currently working on adding examples to the Editor’s draft. A second proposal is currently under review in: https://github.com/w3c/sdw/pull/891


We did indeed take the work in [1] as one of the inputs to our actuation/actuator modelling in SOSA/SSN. We did not include the Effect class, since for W3C recommendations we need implementation evidence for all terms. However, the status of the Stimulus class (for the SSO pattern) and the potential inclusion of an Effect class is currently investigated regarding W3C policy and how classes are treated in an ontology that are merely included for the description of a pattern, rather than for actual usage in instances of that ontology. We have raised an issue: https://www.w3.org/2015/spatial/track/issues/187 that relates to this. If the pattern is included, it would, however be in the SSN module (like the Stimulus class) rather than in the simple SOSA core.

Kind regards,
Armin

On 10/5/17, 3:15 am, "Nicolas Seydoux" <nseydoux@irit.fr> wrote:

    Dear all,
    
    I am a user of the SSN ontology as well. I think its transformation is
    for the best, especially regarding the simplification of its core, its
    effective modularization, and the integration of the notion of Actuator
    and Action.
    
    I second Catherine : the use of comprehensive examples are crucial to
    the proper reuse of a vocabulary, and that maybe the example
    https://github.com/w3c/sdw/blob/gh-pages/ssn/rdf/documentation_examples/sosa-core_examples.ttl

    could be declined in an example demonstrating the use of ssn as well,
    with the description of some SystemCapability for instance, as well as
    the integration of results with qudt:Unit and such.
    
    Moreover, as Mr. Rodriguez pointed out, I proposed an ontology, SAN [1],
    as a mirror to SSN for actuation and actuators. It is dependant on DUL
    as well, but I'll soon release an updated version aligned with SOSA.
    However, I would also like to know if ontology design patterns were
    integrated into SOSA, and if they could be explicitely represented in a
    separated module if so, as I did for AAE, presented in [2] (open access
    publication). It would ease their reusability.
    
    Globally, speaking as the developer of applications based on the
    ontology, I'm really happy about the new version of sosa and ssn.
    
    Sincerely yours, and thank you for your feedbacks,
    
    Nicolas Seydoux
    
    [1] https://www.irit.fr/recherches/MELODI/ontologies/SAN

    
    [2] https://hal.archives-ouvertes.fr/hal-01467853v1

    
    -- 
    Nicolas Seydoux, IRIT/LAAS-CNRS
    PhD student - Semantic web and IoT
    IRIT : MELODI team - Semantic web and ontologies
    LAAS : SARA team - Advanced networks and IoT
    Tel : +33(0)5 61 33 69 07

Received on Monday, 15 May 2017 06:31:42 UTC