Re: complementOf -> viewOf: proposed text

Satya,

but we don't have a notion of classes/instances in PROV. I remember this being discussed early on and dismissed. As a result, you 
can't use that framework so you can't make a distinction between the two cases. It may be in our heads, but it's not in the 
language, right?

--Paolo


  On 1/19/12 1:04 AM, Satya Sahoo wrote:
> Hi Stian,
> Your example really helps in fleshing out the multiple issues that are unfortunately being mixed up (in my view), comments are inline:
>
>     entity(customerOnRedChair, [prov:location="the red chair in the cafe"])
>
> This is a "class" definition (defining a category of resources/entities - as used in maths (sets), logic etc.)
>
>     entity(paoloInCafe)
>     entity(stianInCafe)
>
> These may be classes (with individuals corresponding to all the times that paolo or stian are in cafe) or individuals as required 
> by an application.
>
>     entity(paolo)
>     entity(stian)
>
> These are all "individuals" (not a category of resources as "customerOnRedChair") - these are members of the sets/classes.
>
> I believe you mention this distinction between class and instances in one of your later mails.
>
>     specializationOf(paoloInCafe, paolo)
>     specializationOf(stianInCafe, stian)
>
> I think this construct hides many complexities and is incorrect according well-defined specialization-generalization relationship 
> in logic, programming languages, maths etc. For example a person is a not a specialization of the same person in different 
> situations (and neither are the descriptions/records about that person) :)
>
> I agree with James that many of the examples discussed before your mail were referring to attributes descriptions and not entities.
>
> There are two ways of interpreting the above assertions:
> a) when paoloInCafe is a class (described above) - then the above construct is incorrect as it is mixing "types" (asserting class 
> to be specialization of an individual)
>
> b) when paoloInCafe is an individual - then again the above construct is incorrect as specialization is asserted between classes 
> and not individuals
>
>     alternateOf(paoloInCafe, customerOnRedChair)
>     alternateOf(stianInCafe, customerOnRedChair)
>
>
>     but we probably don't want to then infer:
>     alternateOf(paoloInCafe, stianInCafe)
>
>     and certainly not:
>     alternateOf(paolo, stian)
>
> The above mentioned mixing of types paves the way the following (seemingly) incorrect inference. I disagree with Paolo that the 
> incorrect inference is due to absence of time from the above examples.
>
> Thanks.
>
> Best,
> Satya
>
>     .. neither did overlap the old characterisation intervals, and are
>     different 'things' in the world.
>
>
>     however, if Paolo and Stian did not sit anywhere else but in the red
>     chair, we can also have:
>
>
>     specializationOf(paoloInCafe,
>     customerOnRedChair)specializationOf(stianInCafe, customerOnRedChair)
>     this implies that for the duration of paoloInCafe, it was also
>     customerOnRedChair.
>
>     --
>     Stian Soiland-Reyes, myGrid team
>     School of Computer Science
>     The University of Manchester
>
>


-- 
-----------  ~oo~  --------------
Paolo Missier - Paolo.Missier@newcastle.ac.uk, pmissier@acm.org
School of Computing Science, Newcastle University,  UK
http://www.cs.ncl.ac.uk/people/Paolo.Missier