Re: PROV-ISSUE-89 (what-entity-attributes): How do we find the attributes of an entity? [Formal Model]

Luc,

On May 7, 2012, at 11:37 AM, Luc Moreau wrote:

> It's a very old issue.

Yes :-)

> I thought it was agreed that all properties (which dont encode relations) of an entity would be regarded as attributes.

Yes, I believe this was the resolution.
I think this issue can be closed.

However, you morphed this issue into a note about "proper" provenance (in March).
So may I ask that you open a new issue for that concern, so that I can close this?

Thanks,
Tim

> 
> Professor Luc Moreau
> Electronics and Computer Science
> University of Southampton
> Southampton SO17 1BJ
> United Kingdom
> 
> 
> On 7 May 2012, at 16:30, "Timothy Lebo" <lebot@rpi.edu> wrote:
> 
>> Luc,
>> 
>> I'm looking to close this issue.
>> 
>> "proper" provenance is not part of the current prov-o WD.
>> 
>> As you know, the building story in prov-o is "starting points, expanded, qualified" and then "collections".
>> 
>> Since the notion of "proper" has evolved since F2F2, could you summarize what aspects you think should still be addressed in prov-o?
>> For me, the notion of proper hinges around specOf, where "improper" asserters assert details on less specific entities than they should.
>> 
>> Thanks,
>> Tim
>> 
>> 
>> On Mar 5, 2012, at 5:14 PM, Luc Moreau wrote:
>> 
>>> Hi Tim,
>>> 
>>> I think that somewhere in the prov-o html document, there should be a section
>>> that talks about "proper" provenance (or whatever its name is), and discusses attributes.
>>> So, as long as we remember to discuss this, I think we can close the issue.
>>> 
>>> Luc
>>> 
>>> On 05/03/12 22:06, Timothy Lebo wrote:
>>>> On Mar 5, 2012, at 4:42 PM, Luc Moreau wrote:
>>>> 
>>>> 
>>>>> Hi Tim,
>>>>> 
>>>>> Indeed, we no longer make the distinction between characterizing and non-characterizing attributes.
>>>>> 
>>>>> In "proper" provenance, attributes are still very important, since they help describe a "partial state".
>>>>> Hence, some constraints exist around attributes:
>>>>> http://dvcs.w3.org/hg/prov/raw-file/default/model/prov-dm-constraints.html#account-and-accountEntity
>>>>> (For instance, see 3rd bullet point).
>>>>> 
>>>> Yes, the 3rd bullet and "It is not permitted to add new attributes to a given entity" in the note.
>>>> 
>>>> 
>>>>> But I think we came to the conclusion that any rdf property for an entity is regarded as an attribute.
>>>>> 
>>>> 
>>>> 
>>>> 
>>>>> Isn't it the answer to this issue?
>>>>> 
>>>> I hope that it is. I am comfortable with the phrasing in the DM.
>>>> Any rdf property outside of the prov namespace that describes and Entity is "fixed".
>>>> (as you noted, we can "use" it again, so that would make a new attribute within the prov namespace but that did not affect the partial state.)
>>>> 
>>>> 
>>>> Does this mean we can close the issue? Do we need the DM to say something further in light of this ISSUE?
>>>> 
>>>> Thanks,
>>>> -Tim
>>>> 
>>>> 
>>>> 
>>>>> Cheers,
>>>>> Luc
>>>>> 
>>>>> 
>>>>> On 05/03/12 19:13, Timothy Lebo wrote:
>>>>> 
>>>>>> Luc,
>>>>>> 
>>>>>> The distinction between characterizing attributes and non-characterizing attributes has faded in the latest versions of the DM.
>>>>>> 
>>>>>> Do you still have concerns about being able to find "frozen" attributes for a given entity?
>>>>>> 
>>>>>> Thanks,
>>>>>> Tim
>>>>>> 
>>>>>> 
>>>>>> On Sep 2, 2011, at 4:52 AM, Provenance Working Group Issue Tracker wrote:
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>>> PROV-ISSUE-89 (what-entity-attributes): How do we find the attributes of an entity? [Formal Model]
>>>>>>> 
>>>>>>> http://www.w3.org/2011/prov/track/issues/89
>>>>>>> 
>>>>>>> Raised by: Luc Moreau
>>>>>>> On product: Formal Model
>>>>>>> 
>>>>>>> The conceptual model defines an entity in terms of an identifier and a list of attribute-value pairs. It is indeed crucial for the asserter to identify the attributes that have been frozen in a given entity.
>>>>>>> 
>>>>>>> Currently, the ontology does not seem to identify these attributes.
>>>>>>> 
>>>>>>> To say that these attributes could be found by looking at all the properties for this entity does not work with an open world assumption.
>>>>>>> 
>>>>>>> What mechanism do we have to identify these attributes?
>>>>>>> 
>>>>>>> 
>>>>>>> 
>>>>>>> 
>>>>>>> 
>>>>>>> 
>>>>>>> 
>>>>>>> 
>>>>>> 
>>>>>> 
>>>>> 
>>>> 
>>> 
>> 
> 
>

Received on Monday, 7 May 2012 15:44:45 UTC