RE: The wasQuotedFrom relationship

Hello Stian,

I'm not sure it's actually wrong to model a blog post as a collection, and if I had to model the relation in my own application (as opposed to needing intuitive examples for the primer), I would probably do so. From what I had understood from the DM, a PROV collection is an entity we want to model as composite and whose components may change over time. The distinction with a Dublin Core resource with parts seems to be that PROV makes it easier to describe the sequence of changes, i.e. starting with an empty collection and changing membership from one specialized version of the collection to the next. I would even consider using prov:Dictionary, as this would then allow me to specify how the blog post and quote were related.

The fact that we are unsure or disagree about the scope of use of prov:Collection might suggest that its use should be clarified in the primer and/or FAQ (unless it is just that I've misunderstood and everyone else in the WG thinks it's clear).

thanks,
Simon

Dr Simon Miles
Senior Lecturer, Department of Informatics
Kings College London, WC2R 2LS, UK
+44 (0)20 7848 1166

Mapping Dublin Core (Attribution Metadata) to the Open Provenance Model:
http://eprints.dcs.kcl.ac.uk/1386/
________________________________
From: stian@mygrid.org.uk [stian@mygrid.org.uk] on behalf of Stian Soiland-Reyes [soiland-reyes@cs.manchester.ac.uk]
Sent: 20 February 2013 10:29
To: Miles, Simon
Cc: Timothy Lebo; public-prov-wg@w3.org; pgroth@gmail.com
Subject: RE: The wasQuotedFrom relationship


Agree that hadMember would not be good, making a blog post a collection is quite confusing,  if not wrong.

If we need to say something, we should just relate them with dcterms:hasPart, as we (perhaps sadly)
decided to not cover entity partOf entity in PROV.

However I think for the primer we are fine unless someone outside asks for that relation.

--
Stian Soiland-Reyes
myGrid team, University of Manchester
http://soiland-reyes.com/stian/work

On 20 Feb 2013 00:02, "Miles, Simon" <simon.miles@kcl.ac.uk<mailto:simon.miles@kcl.ac.uk>> wrote:
Stian,

Yes, I also thought about expressing the containment relation between the blog post and quote. I noticed that you and Tim used dcterms:hasPart to express this in this mail thread. We also have prov:hadMember, which might be more appropriate, as the fact that the quote is part of the blog entry is a possibly temporary past state (implied in the primer example by the fact that the article it quotes from is updated), which dcterms:hasPart doesn't obviously capture. Also, we recommend dcterms:hasPart for relating PROV activities, whereas these are entities. On the other hand, using prov:hadMember would make the blog entry a prov:Collection, and it is not the most intuitive example of a collection for a primer.

In the end, the implications seemed too complicated for a primer, especially as the blog entry entity is not itself used anywhere else in the example, so I left it out. There might be an intuitive, succinct and unambiguous way to introduce it, though, if we thought it useful.

thanks,
Simon

Dr Simon Miles
Senior Lecturer, Department of Informatics
Kings College London, WC2R 2LS, UK
+44 (0)20 7848 1166<tel:%2B44%20%280%2920%207848%201166>

Mapping Dublin Core (Attribution Metadata) to the Open Provenance Model:
http://eprints.dcs.kcl.ac.uk/1386/

________________________________________
From: stian@mygrid.org.uk<mailto:stian@mygrid.org.uk> [stian@mygrid.org.uk<mailto:stian@mygrid.org.uk>] on behalf of Stian Soiland-Reyes [soiland-reyes@cs.manchester.ac.uk<mailto:soiland-reyes@cs.manchester.ac.uk>]
Sent: 19 February 2013 23:29
To: Miles, Simon
Cc: pgroth@gmail.com<mailto:pgroth@gmail.com>; Timothy Lebo; public-prov-wg@w3.org<mailto:public-prov-wg@w3.org> Group WG
Subject: Re: The wasQuotedFrom relationship

This reads well in the primer and in your response. The combination of
ex:quoteInBlogEntry and prov:value here makes it quite obvious.  If we
want to expand it more we could use html blockquote, id and RDFa
argument.

An open question could be how we know that ex:quoteInBlogEntry is part
of ( ex:blogPost ?)  , but as we just skim and don't mention the blog
post I think we can get away with the current text. :-)   (It is kind
of out of scope of PROV to define such kind of containment or
belonging).





On Tue, Feb 19, 2013 at 6:32 PM, Miles, Simon <simon.miles@kcl.ac.uk<mailto:simon.miles@kcl.ac.uk>> wrote:
> Tim, Paul, Stian, all,
>
> It is clear that, to resolve the issue discussed below, a "quote in blog
> entry" entity needs to be introduced into the primer. I've constructed a
> response below, based on your feedback. Please also the revised primer,
> start of Section 3.9.
>
> http://dvcs.w3.org/hg/prov/raw-file/default/primer/Primer.html
>
> Does this seem an adequate response to Chuck?
>
> thanks,
> Simon
>
> ===
> After discussion, we agree with you that the PROV primer was still unclear,
> or possibly just wrong, in the way it was implying wasQuotedFrom could be
> used. As you say, one would not say that "X was quoted from Y" if X was not
> a quotation. We still believe the relation itself, as defined in the PROV
> specifications, is correct and unambiguous.
>
> We have revised the primer again following your suggestion of introducing an
> entity that is more clearly a quotation, ex:quoteInBlogEntry, and made
> explicit the text actually quoted ("Smaller cities have more crime than
> larger ones.")
>
> With regards to wasQuotedFrom itself, we note that "X wasQuotedFrom Y"
> implies that X is a quotation, and that this follows the same idea of
> quotation as in HTML ("The blockquote element represents a section that is
> quoted from another source", HTML5). PROV does not provide a relation "X was
> quoted from in Y".
>
> Please see the revised primer at the link below. The relevant text and
> example are at the start of Section 3.9, as before.
>
> http://dvcs.w3.org/hg/prov/raw-file/default/primer/Primer.html
>
> Do you believe this now addresses your concern?
> ===
>
>
> From: Paul Groth [pgroth@gmail.com<mailto:pgroth@gmail.com>]
> Sent: 11 February 2013 20:50
> To: Timothy Lebo
> Cc: Stian Soiland-Reyes; Miles, Simon; public-prov-wg@w3.org<mailto:public-prov-wg@w3.org> Group WG
> Subject: Re: The wasQuotedFrom relationship
>
> Oh just saw that html5 defines blockquote as:
>
> "The blockquote element represents a section that is quoted from another
> source"
>
> I think prov:wasQuotedFrom fits that definition perfectly.
>
> cheers
> Paul
>
> P.S. We should write a blog post about how to use prov with html5
>
>
>
>
>
> On Mon, Feb 11, 2013 at 8:17 PM, Timothy Lebo <lebot@rpi.edu<mailto:lebot@rpi.edu>> wrote:
>>
>>
>> On Feb 11, 2013, at 11:27 AM, Stian Soiland-Reyes
>> <soiland-reyes@cs.manchester.ac.uk<mailto:soiland-reyes@cs.manchester.ac.uk>> wrote:
>>
>> PROV can cover a lot of things, but I just hope we have not just made
>> a kind of "SGML of provenance" in that it allows anything and
>> recommends nothing, as then you are still just as confused after
>> reading the specs, and as a result everybody would end up using PROV
>> differently.
>>
>>
>> Yes, there's a risk that if we under specify that many will use it
>> differently. But the WG is simply providing the core.
>> As long as people are conforming to Activity and Entity, we should be
>> okay…
>>
>> -Tim
>>
>



--
Stian Soiland-Reyes, myGrid team
School of Computer Science
The University of Manchester