W3C home > Mailing lists > Public > public-semweb-lifesci@w3.org > June 2014

Re: License unknown

From: Alan Ruttenberg <alanruttenberg@gmail.com>
Date: Wed, 18 Jun 2014 14:06:53 -0400
Message-ID: <CAFKQJ8=b9T9SafpJcLsA9NRxfkWMXQUnsmOv_V2fEKRnaBe1pg@mail.gmail.com>
To: "M. Scott Marshall" <mscottmarshall@gmail.com>
Cc: "Eric Prud'hommeaux" <eric@w3.org>, "Gray, Alasdair J G" <A.J.G.Gray@hw.ac.uk>, Jerven Bolleman <me@jerven.eu>, HCLS IG <public-semweb-lifesci@w3.org>
Yes, mostly. I'm also trying to adhere to the ideal of having the semweb
assertions being a monotonically increasing set. At the moment the
semantics of the reasoning tools assume monotonicity. For an example of the
problem, consider a resource which makes inferences based on assertions
other make. Keeping their resource up-to-date is much harder if they have
to keep checking if assertions have disappeared, and then run their
reasoning process again from the start.

-Alan


On Wed, Jun 18, 2014 at 2:01 PM, M. Scott Marshall <mscottmarshall@gmail.com
> wrote:

> Ah, got it. I was creating the same problem that you were attempting
> to fix: You want to make it impossible to add an inconsistent
> assertion in the case that the license is initially unknown but is
> determined later by building the safe guard into the property. That
> seems reasonable enough.. although we (Alasdair and maybe more?) are
> creating a validator, as well as an editor (building on the VOID
> editor) which would prevent such inconsistencies anyway. Still, it
> solves the problem of being able to require that data publishers
> actually *try* to determine the license in order to meet our
> specification without going against best (?) 'open world' practices,
> where unstated = unknown. It also avoids the problem of strictly
> limiting datasets to 1 license - probably too restrictive.
>
> A small doubt: it's still possible to create an inconsistency when
> :tried-to-determine-license "false"^xsd:Boolean but dc:license has a
> valid license as value.
>
> Alasdair, Eric - Is there anything in the proposed solution that
> http://www.w3.org/2013/ShEx/Primer cannot catch?
>
> Jerven - does it handle your objections?
>
> -Scott
>
>
> On Wed, Jun 18, 2014 at 3:06 PM, Alan Ruttenberg
> <alanruttenberg@gmail.com> wrote:
> > The reason I labeled it tried-to-determine-license is that that way the
> > assertion remains true. If the predicate is whether we *have* determined
> > license then the assertion needs to be retracted when we do. When
> possible
> > we try to make statements that remain true.
> >
> > When probing for a license, first check dc:license, and if it is absent,
> > check :tried-to-determine-license.
> >
> > In user interfaces there's no reason to show both. It's perfectly
> reasonable
> > to use these properties to decide whether to show "License: Unknown"
> >
> > -Alan
> >
> >
> > On Wed, Jun 18, 2014 at 8:46 AM, M. Scott Marshall
> > <mscottmarshall@gmail.com> wrote:
> >>
> >> Just a slight tweak to Alan's suggestion (thanks Alan):
> >> :determined-license "true"^xsd:Boolean
> >>
> >> -Scott
> >>
> >> On Wed, Jun 18, 2014 at 2:36 PM, Alan Ruttenberg
> >> <alanruttenberg@gmail.com> wrote:
> >> > I concur. This is clear from the semantics of OWL. Understand the
> >> > difference
> >> > between integrity constraints and OWL assertions.
> >> >
> >> > If you want to say that there was an attempt to find a license and
> that
> >> > it
> >> > couldn't be found, say that. You could do so as an annotation on the
> >> > axiom,
> >> > or as a distinct property e.g.
> >> > :tried-to-determine-license "true"^xsd:Boolean
> >> >
> >> > This can remain true, as a historical fact, even after a license is
> >> > found.
> >> >
> >> > For more about integrity constraints see
> >> > http://clarkparsia.com/pellet/icv/
> >> >
> >> > -Alan
> >> >
> >> >
> >> >
> >> > On Monday, June 16, 2014, Jerven Bolleman <me@jerven.eu> wrote:
> >> >>
> >> >> Hi Alasdair,
> >> >>
> >> >> I think you are closing the world the wrong way. Both in limiting a
> >> >> dataset to have one license and secondly by having an "unknown"
> string
> >> >> to encode that you don't know something.
> >> >> Instead of using "unknown" you should use a blank node.
> >> >>
> >> >> cc:CC0 owl:sameAs "unknown"
> >> >>
> >> >> is plain wrong and can't be inferred
> >> >>
> >> >> cc:CC0 owl:sameAs _:1
> >> >>
> >> >> is fine and a nice way of allowing dataset from different sources
> >> >> descriptions to be combined and give more information.
> >> >>
> >> >> If you don't want a blank node, make a local node instead.
> >> >> cc:CC0 owl:sameAs <./#license>
> >> >>
> >> >> I implore you to change your mind on this. Choose semantics over
> >> >> special
> >> >> values!
> >> >>
> >> >> Regards,
> >> >> Jerven
> >> >>
> >> >>
> >> >> On Mon, Jun 16, 2014 at 5:44 PM, Gray, Alasdair J G
> >> >> <A.J.G.Gray@hw.ac.uk>
> >> >> wrote:
> >> >> > Hi Jerven,
> >> >> >
> >> >> > We are trying to ensure that a license is provided for a dataset,
> and
> >> >> > that
> >> >> > only one license is provided. (We are closing the world in certain
> >> >> > areas.)
> >> >> >
> >> >> > Our proposed solution is written up at
> >> >> > https://github.com/joejimbo/HCLSDatasetDescriptions/issues/65
> >> >> >
> >> >> > Alasdair
> >> >> >
> >> >> > On 2 Jun 2014, at 16:47, Jerven Bolleman <me@jerven.eu> wrote:
> >> >> >
> >> >> > Hi Alasdair,
> >> >> >
> >> >> > It is good practice in my opinion to not state unknown facts. If
> you
> >> >> > want to state there should be a license but you don't know it I
> think
> >> >> > some owl modelling is called for.
> >> >> >
> >> >> > e.g. roughly like this.
> >> >> >
> >> >> > dusty_data a dct:Dataset .
> >> >> > dusty_data rdfs:subClassOf [ a owl:Restriction ;
> >> >> >                                            owl:onProperty
> dct:license
> >> >> > ;
> >> >> >                                            owl:minCardinality 1 ]
> >> >> >
> >> >> >
> >> >> > If anyone does something like
> >> >> >
> >> >> > dusty_data a dct:dataset .
> >> >> > dusty_data dct:license bad_example:unknown_license .
> >> >> >
> >> >> > and then someone figures out it has a license .
> >> >> >
> >> >> > dusty_data a dct:Dataset .
> >> >> > dusty_data dct:license cc:zero .
> >> >> >
> >> >> > And puts all those triples into one store/ontology then queries
> will
> >> >> > start to return wrong data.
> >> >> >
> >> >> > SELECT ?datasetWithALicense
> >> >> > WHERE {
> >> >> > ?datasetWithALicense dct:license ?license .
> >> >> > }
> >> >> >
> >> >> > Will give 2 bindings for ?dataset and ?license one of them being
> >> >> > wrong
> >> >> > now.
> >> >> >
> >> >> > If you ask the same question depending on the better modelling then
> >> >> > you get the correct answer in both cases.
> >> >> >
> >> >> > SELECT ?datasetWIthoutLicense
> >> >> > WHERE {
> >> >> > ?datasetWIthoutLicense a dct;Dataset .
> >> >> > MINUS ( ?datasetWIthoutLicense dct:license ?license )
> >> >> > }
> >> >> >
> >> >> > SELECT ?datasetWIthALicense
> >> >> > WHERE {
> >> >> > ?datasetWIthALicense a dct;Dataset .
> >> >> > ?datasetWIthALicense dct:license ?license .
> >> >> > }
> >> >> >
> >> >> > Both give the expected results.
> >> >> >
> >> >> >
> >> >> > In general in RDF just state what you know and describe what you
> >> >> > don't
> >> >> > know.
> >> >> >
> >> >> > Regards,
> >> >> > Jerven
> >> >> >
> >> >> > On Mon, Jun 2, 2014 at 5:25 PM, Gray, Alasdair J G
> >> >> > <A.J.G.Gray@hw.ac.uk>
> >> >> > wrote:
> >> >> >
> >> >> > Hi,
> >> >> >
> >> >> > I am working to help shape the Healthcare and Life Sciences
> community
> >> >> > profile for describing datasets, current version available from [1]
> >> >> >
> >> >> > One of our goals is that there is a minimal set of properties that
> >> >> > are
> >> >> > available for all datasets, and we would like these properties to
> >> >> > include
> >> >> > the license. However we have the problem that for several legacy
> >> >> > datasets
> >> >> > the license is simply unknown. Does anyone know of a resource that
> >> >> > can
> >> >> > be
> >> >> > used to represent that the license is unknown as the value of a
> >> >> > dcterms:license predicate?
> >> >> >
> >> >> > Thanks
> >> >> >
> >> >> > Alasdair
> >> >> >
> >> >> > [1] https://github.com/joejimbo/HCLSDatasetDescriptions
> >> >> >
> >> >> > Alasdair J G Gray
> >> >> > Lecturer in Computer Science, Heriot-Watt University, UK.
> >> >> > Email: A.J.G.Gray@hw.ac.uk
> >> >> > Web: MailScanner has detected a possible fraud attempt from
> >> >> > "www.macs.hw.ac.uk" claiming to be http://www.alasdairjggray.co.uk
> >> >> > ORCID: http://orcid.org/0000-0002-5711-4872
> >> >> > Telephone: +44 131 451 3429
> >> >> > Twitter: @gray_alasdair
> >> >> >
> >> >> >
> >> >> >
> >> >> >
> >> >> >
> >> >> >
> >> >> > ________________________________
> >> >> >
> >> >> > Sunday Times Scottish University of the Year 2011-2013
> >> >> > Top in the UK for student experience
> >> >> > Fourth university in the UK and top in Scotland (National Student
> >> >> > Survey
> >> >> > 2012)
> >> >> >
> >> >> > We invite research leaders and ambitious early career researchers
> to
> >> >> > join us
> >> >> > in leading and driving research in key inter-disciplinary themes.
> >> >> > Please
> >> >> > see
> >> >> > www.hw.ac.uk/researchleaders for further information and how to
> >> >> > apply.
> >> >> >
> >> >> > Heriot-Watt University is a Scottish charity registered under
> charity
> >> >> > number
> >> >> > SC000278.
> >> >> >
> >> >> >
> >> >> >
> >> >> >
> >> >> > --
> >> >> > Jerven Bolleman
> >> >> > me@jerven.eu
> >> >> >
> >> >> >
> >> >> > Alasdair J G Gray
> >> >> > Lecturer in Computer Science, Heriot-Watt University, UK.
> >> >> > Email: A.J.G.Gray@hw.ac.uk
> >> >> > Web: MailScanner has detected a possible fraud attempt from
> >> >> > "www.macs.hw.ac.uk" claiming to be http://www.alasdairjggray.co.uk
> >> >> > ORCID: http://orcid.org/0000-0002-5711-4872
> >> >> > Telephone: +44 131 451 3429
> >> >> > Twitter: @gray_alasdair
> >> >> >
> >> >> >
> >> >> >
> >> >> >
> >> >> >
> >> >> >
> >> >> > ________________________________
> >> >> >
> >> >> > Sunday Times Scottish University of the Year 2011-2013
> >> >> > Top in the UK for student experience
> >> >> > Fourth university in the UK and top in Scotland (National Student
> >> >> > Survey
> >> >> > 2012)
> >> >> >
> >> >> > We invite research leaders and ambitious early career researchers
> to
> >> >> > join us
> >> >> > in leading and driving research in key inter-disciplinary themes.
> >> >> > Please
> >> >> > see
> >> >> > www.hw.ac.uk/researchleaders for further information and how to
> >> >> > apply.
> >> >> >
> >> >> > Heriot-Watt University is a Scottish charity registered under
> charity
> >> >> > number
> >> >> > SC000278.
> >> >>
> >> >>
> >> >>
> >> >> --
> >> >> Jerven Bolleman
> >> >> me@jerven.eu
> >> >>
> >>
> >> --
> >> M. Scott Marshall, PhD
> >> MAASTRO clinic, http://www.maastro.nl/en/1/
> >> http://eurecaproject.eu/
> >> http://semantic-dicom.org/
> >> https://plus.google.com/u/0/114642613065018821852/posts
> >> http://www.linkedin.com/pub/m-scott-marshall/5/464/a22
>
Received on Wednesday, 18 June 2014 18:07:53 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 14:53:09 UTC