W3C home > Mailing lists > Public > public-xg-prov@w3.org > September 2010

Re: "oh yeah?" - button

From: Paul Groth <pgroth@gmail.com>
Date: Wed, 08 Sep 2010 11:45:14 +0200
Message-ID: <4C875B2A.3050700@gmail.com>
To: Reto Bachmann-Gmuer <reto@apache.org>
CC: public-xg-prov@w3.org
Hi Reto,

I think the notion of multiple parties having different views of a 
shared world is captured within the use cases. I think most notably in 
the business contract scenario where two parties have different views on 
the same business transaction.

While I believe what you are mentioning is covered by the requirements, 
it is perhaps not emphasized because the use cases gathered (over 30) 
emphasized the problem of being able to determine the provenance of an 
object across multiple independent systems.

Also, note, that the literature on provenance does cater for your notion 
of justifying beliefs particularly from the intelligence domain (see 
work from Gil namely, Trellis, Inference Web people namely IWTrust).

Paul



Reto Bachmann-Gmuer wrote:
> Thanks Paul for the pointers.
>
> Both the concept of content as well as the seal seem to be based on
> the notion of original content by a single source, and metadata on
> which entailment and inference is to be applied. When it comes to
> facts about a shared world multiple parties may come to assert the
> same triples independently. In this case the inference is not
> something applied on the provenance record but to the believes that
> are the object of provenance. Alice has reason to believe that there
> is something that is a cat and can fly because:
>
> - Bob says that :Felix can fly and Susanne says that :Felix is a cat
> - John says that tigers can fly and Peter says that tigers are cats
>
> I think the provenace of artifacts and contents which the use cases
> seem to focus on is an application of the more general question on the
> provenance/justification of believes.
>
> Cheers,
> Reto
>
> On Tue, Sep 7, 2010 at 4:39 PM, Paul Groth<pgroth@gmail.com>  wrote:
>    
>> Hi Reto,
>>
>> If you take a look at the discussion of justifications for decisions in the
>> Content section of the User Requirements document
>> (http://www.w3.org/2005/Incubator/prov/wiki/User_Requirements#Content).
>> There's a notion of using provenance to justify why a system presents a
>> particular outcome, in your case Why something is both a cat and a flying
>> object.
>>
>> Also, in the News Aggregator Scenario, there's a notion of a seal (similar
>> to the "oh yeah button") where users can find out why a post was constructed
>> in a particular fashion.
>>
>> I hope that clarifies things.
>>
>> Paul
>>
>> Reto Bachmann-Gmuer wrote:
>>      
>>> Hello
>>>
>>> I was just having a first read of the use cases and requirements. Most
>>> use cases seem to assume single provenances of assertion and I was
>>> wondering where a "oh yeah?"-button would fit in.
>>>
>>> Something Alices presses on when she reads [ rdf:type eg:cat; rdf:type
>>> eg:flyingObject] to get an enumeration of sets  of (sub)graphs and
>>> their (somehow trusted) source that entail that there is something
>>> that flies and is a cat.
>>>
>>> Cheers,
>>> reto
>>>
>>>
>>>
>>>        
Received on Wednesday, 8 September 2010 09:50:20 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 8 September 2010 09:50:21 GMT