Re: License unknown

On Mon, Jun 23, 2014 at 11:27 AM, Jerven Bolleman <me@jerven.eu> wrote:

> Again, I think we should encourage more specificity than the boolean
> property flag.
>
> One academic makes the following statements
>
> _:sillyAcademicsDataset a dct:Dataset ;
>                   prov:wasDerivedFrom realworld:commercialDataset .
> realworld:commercialDataset a dct:Dataset ;
>                   :tried-to-determine-license true .
>
> Second Academic makes these statements
> _:lazyAcademicsDataset a dct:Dataset ;
>                   prov:wasDerivedFrom realworld:commercialDataset .
> realworld:commercialDataset a dct:Dataset ;
>                   :tried-to-determine-license true .
>
> Combine this knowledge
> realworld:commercialDataset a dct:Dataset ;
>                   :tried-to-determine-license true ;
>                   :tried-to-determine-license false .
>

This is a better criticism. While this can happen, I'd suggest that the
false assertion could not be asserted correctly, for the practical reason
that a single researcher can not know if there never has been a case of
trying to determine license.  However I would add a note of documentation
(rdfs:comment is perfect for this) to this effect.
Once we are dealing with asserting incorrect facts, anything goes. (I
didn't say the property value was whether "*" tried to determine it.
However if I tried to determine it then I know that a case of trying has
happened. There's a reason that it isn't good practice to use unqualified
pronouns on the web)


> Not very helpful.
>
> Take the rdfs:comment approach
> _:sillyAcademicsDataset a dct:Dataset ;
>                   prov:wasDerivedFrom realworld:commercialDataset .
> realworld:commercialDataset a dct:Dataset ;
>                   dct:license [rdfs:comment "I tried to look at XXX,
> and believe its fair use because of YYYY" ]
>
> _:lazyAcademicsDataset a dct:Dataset ;
>                   prov:wasDerivedFrom realworld:commercialDataset .
> realworld:commercialDataset a dct:Dataset ;
>                   dct:license [rdfs:comment "Lazy academic did not
> care to look for the license"]
>
> Combine this knowlege
> realworld:commercialDataset a dct:Dataset ;
>                   dct:license [rdfs:comment "I tried to look at XXX,
> and believe its fair use because of YYYY" ],
>                                    [rdfs:comment "Lazy academic did
> not care to look for the license"] .
>
> Its still use full.
>
> Then try this
> _:sillyAcademicsDataset a dct:Dataset ;
>                   prov:wasDerivedFrom realworld:commercialDataset .
> realworld:commercialDataset a dct:Dataset ;
>                   dct:license [a :UseAtOwnLegalRisk .
>                                     rdfs:comment "I tried to look at
> XXX, and believe its fair use because of YYYY" ]
>
> Even more rich information an end user can actually use.
>

A user *reading* the comment can use it. A machine (an essential component,
ideally, of our target audience) will have a harder time.


> One has to think about more than one re-publisher of a dataset, as
> that does happen in our field (see drugbank as a real life example).
>

Yes. And republishers need to think more carefully about the consequences
of republishing copyrightable information without knowing the license terms.

-Alan


>
>
> Regards,
> Jerven
>
>
>
> On Mon, Jun 23, 2014 at 5:17 PM, Alan Ruttenberg
> <alanruttenberg@gmail.com> wrote:
> >
> >
> >
> > On Mon, Jun 23, 2014 at 11:08 AM, Joachim Baran <joachim.baran@gmail.com
> >
> > wrote:
> >>
> >>
> >> On 23 June 2014 06:37, Alan Ruttenberg <alanruttenberg@gmail.com>
> wrote:
> >>>
> >>> In the case that the license is not asserted it distinguishes the case
> >>> where the publisher has made an affirmative effort to determine what
> the
> >>> license is, or not.
> >>
> >>   I cannot fathom how this could be inferred from the truth value of a
> >> bit.
> >>
> >
> > By documenting the property so as to make that explicit. How else do we
> > understand any property value?
> >
> > -Alan
> >
>
>
>
> --
> Jerven Bolleman
> me@jerven.eu
>