RE: Alignment between prov-o and SWEET

Well, PROV is not planned for any updates (beyond bugfixes) at the moment, but you are free to use PROV while violating some of the PROV Constraints http://www.w3.org/TR/prov-constraints/ -  at the loss of not being able to make reliable logical inferences from it. (That is – use PROV-O as a Vocabulary rather than an Ontology – meaning in OWL you should then not use owl:import).

You may find that aligning your ontology to PROV requires you to rethink some of your own concepts, as for instance a PROV Entity should have a consistent Generation, and you have to watch out for any assumptions that made sense in your ontology in isolation, but which now might imply (in PROV Semantics) impossibilities or inconsistencies.

Say for instance you have a Document and a Curator, which you now map to PROV and get (after OWL inference) an Entity of original document that wasGeneratedBy an Activity that started and ended in 1860, which is also somehow wasAttributedTo a modern day curator Agent. The agent might have participated in another Activity started in 2017 – meaning from PVOC constraints that this curator must have existed both in 1860 and 2017. If you end up expressing this formally, it sometimes means splitting entities/activities and relating them using PROV specializations, derivation and activity interactions.  (expressing “Back To the Future” character’s provenance in PROV left as exercise to the reader)

I guess these kind of considerations would apply for any case of retrofitting an upper ontology - however, looking at SWEET I don’t think you would have many of those issues to deal with. 😊

--
Stian Soiland-Reyes, eScience Lab
School of Computer Science, The University of Manchester
http://orcid.org/0000-0001-9842-9718


From: lewis john mcgibbney<mailto:lewismc@apache.org>
Sent: 05 September 2017 19:21
To: Stian Soiland-Reyes<mailto:soiland-reyes@manchester.ac.uk>
Cc: public-prov-comments@w3.org<mailto:public-prov-comments@w3.org>
Subject: Re: Alignment between prov-o and SWEET

Hello Stian,
Thanks for your comment here and participation over on the SWEET Github issue tracker.
Can you please clarify the following. You've stated "...it might mean you having to challenge some of the existing ontology design..." does this relate to PROV, SWEET or both?
Thank you again Stian for correcting the hyperlink.
Lewis

On Mon, Sep 4, 2017 at 9:34 AM, Stian Soiland-Reyes <soiland-reyes@manchester.ac.uk<mailto:soiland-reyes@manchester.ac.uk>> wrote:
I would commend your alignment towards PROV 😊 – although it might mean you having to challenge some of the existing ontology design, at least if you want to go for a “sound” (semantically consistent) integration rather than a quicker “scruffy” provenance.

Just a note – I think you sent the wrong link. I found the PROV discussion at https://github.com/ESIPFed/sweet/issues/28


--
Stian Soiland-Reyes, eScience Lab
School of Computer Science, The University of Manchester
http://orcid.org/0000-0001-9842-9718


From: lewis john mcgibbney<mailto:lewismc@apache.org>
Sent: 31 August 2017 15:15
To: public-prov-comments@w3.org<mailto:public-prov-comments@w3.org>
Subject: Alignment between prov-o and SWEET

Hi Folks,
We recently open sourced and transitioned the SWEET Ontology suite [0] from NASA JPL over to the ESIP Federation [1] and have renewed the interest in the resource from within ESIP and further afield. I'm currently spearheading an effort to bring SWEET up-to-date with a wide variety of advances in ontologies such as the excellent work undertaken by this WG on prov-o.
Right now I am working on aligning SWEET with prov-o and would really appreciate input/peer review from this WG on alignment accuracy and extent. So far initial discussion can be seen at [2].
I propose to use the Alignment API software [3] developed by INRIA Rhône-Alpes, France to determine basic alignments. My choice of alignment software was determined by reviewing a larger list of ontology matching software [4] and making a judgement about software usability in order to determine instances of owl:equivalentClass.
As you can see, very little work has been done. I am therefore looking to extend alignments in any way, shape or form which is accurate and factually correct. If you are able to provide input on wherever else I may be missing additional alignments, please comment at [2].
Thank you kindly in advance,
Lewis

[0] https://github.com/ESIPFed/sweet

[1] http://esipfed.org

[2] https://github.com/ESIPFed/sweet/pull/38

[3] http://alignapi.gforge.inria.fr/align.html

[4] http://ontologymatching.org/projects.html


--
http://home.apache.org/~lewismc/

@hectorMcSpector
http://www.linkedin.com/in/lmcgibbney




--
http://home.apache.org/~lewismc/

@hectorMcSpector
http://www.linkedin.com/in/lmcgibbney