Re: PROV-ISSUE-515: Data Model Section 5.1.8 [prov-dm]

Dear all,

I have drafted a response to the following issue at:
http://www.w3.org/2011/prov/wiki/ResponsesToPublicComments#ISSUE-515_.28Invalidation.29

Feedback appreciated,
Regards,
Luc


      ISSUE-515 (Invalidation)

    * Original email:
      http://lists.w3.org/Archives/Public/public-prov-wg/2012Sep/0105.html
    * Tracker: http://www.w3.org/2011/prov/track/issues/515
    * Group Response
          o "Is it possible for entities to become temporarily
            unavailable (e.g., for usage)? "
                + prov-constraints states that usage precedes
                  invalidation and follows generation. Hence, for a
                  given entity, one cannot express that usage is not
                  permitted over a period of time.
                + Alternatively, one can introduce multiple
                  specializations for the various intervals. In the
                  following example, one defines an entity e, and two
                  specializations. e1 is not available after 10am, and
                  e2 before 4pm on 2011-11-16. Both e1 and e2 have an
                  attribute ex:available indicating their availability.
                  On the other hand, e does not have such attribute,
                  because this aspect is not fixed during the lifetime of e.

entity(e)
entity(e1, [ex:available="yes"])
wasInvalidatedBy(e1,-,2011-11-16T10:00:00)
specializationOf(e1,e)
entity(e2,  [ex:available="yes"])
wasGeneratedBy(e2,-,2011-11-16T16:00:00)
specializationOf(e2,e)
wasDerivedFrom(e2,e1)

    *
          o The above example shows that e has some aspects that remain
            constant during its lifetime (e.g. its identity), but is
            also allowed to have other aspects that change over time.
            These changing aspects cannot be expressed as attributes.
          o There is no requirement for asserters to assert invalidation
            of entities
          o Given this, the Working Group feels that the concern raised
            by the author is not applicable. Entities may have long
            lifespan, provided that they have some aspects, represented
            as attributes, that do not change over that lifespan. Other
            aspects are allowed to change. As a minimum, an entity must
            have a fixed entity during its lifetime.
          o As far as a new section on state is concerned, the Working
            Group has made a decision to leave this kind of material
            outside the prov-dm document. Some of this is actually
            covered in prov-constraints.
    * References:
          o Constraint 39:
            http://www.w3.org/TR/2012/WD-prov-constraints-20120911/#generation-precedes-usage
          o Constraint 40:
            http://www.w3.org/TR/2012/WD-prov-constraints-20120911/#usage-precedes-invalidation
          o Section on entities:
            http://www.w3.org/TR/2012/WD-prov-constraints-20120911/#entities--activities-and-agents
          o Group resolution on restructuring prov-dm:
            http://www.w3.org/2011/prov/meeting/2012-02-23#resolution_2
          o Discussion thread on this issue:
    * Implemented Changes to the document:
      http://dvcs.w3.org/hg/prov/rev/f0e8bc2ae457
    * Original author's acknowledgement:


    [edit
    <http://www.w3.org/2011/prov/wiki/index.php?title=ResponsesToPublicComments&action=edit&section=36>]



On 10/09/2012 09:44, Provenance Working Group Issue Tracker wrote:
> PROV-ISSUE-515: Data Model Section 5.1.8   [prov-dm]
>
> http://www.w3.org/2011/prov/track/issues/515
>
> Raised by: Luc Moreau
> On product: prov-dm
>
>
> http://www.w3.org/2011/prov/wiki/LC_Feedback#Data_Model_Section_5.1.8
>
> ISSUE-463
>
> See comments for 5.1.3 (incorporated by reference).
>
> Typo: "in the last three cases" should be "in the third case".
>
> Invalidating an entity due to a state change is going to be difficult for some people to accept, even though it may not be strictly accurate from a rigorous philosophical point of view (not all adopters of PROV will use the model this way). In fact, if this definition is applied consistently throughout the spec, then all entities will have infinitesimally short lifespans and examples (such as the car relocation example in 2.1.1) will become extremely complex. It might be worthwhile adding a section to discuss topics related to entity state, creation, etc. This would provide a way to retain these more complex points while simplifying the examples used throughout the rest of the spec.
>
> Is it possible for entities to become temporarily unavailable (e.g., for usage)? If so, a state model for entities might be helpful.
>
>
>
>
>    

-- 
Professor Luc Moreau
Electronics and Computer Science   tel:   +44 23 8059 4487
University of Southampton          fax:   +44 23 8059 2865
Southampton SO17 1BJ               email: l.moreau@ecs.soton.ac.uk
United Kingdom                     http://www.ecs.soton.ac.uk/~lavm