Re: ISSUE-18: What are the target entities we want relate people to? [People]

> Are you sure this is in scope, Michael? You can describe a person in  
> two distinct ways:

No, I'm not sure. Are you sure it's not in scope?

Have you been present at the last call? If not, pls have a look at the  
minutes [1].

For convenience, I repeat what I said:

1. Our charter only says 'People, such as elements of FOAF or vCard in  
RDF. This is an area for particular attention to privacy  
considerations.' - this doesn't tell us really anything what exactly  
we should do.

2. I suggested that we DO NOT start the discussion around 'this vocab  
should be used or this one should not be used'. Sorry if I gave the  
impression, I just wanted to show how it potentially could look like  
so that people have a rough idea - just pretend sec 4 - 6 doesn't  
exist, for now ;)

3. I suggested that we DO start with use cases (either something  
people wanna do or derived from real-world apps/datasets) and then  
derive requirements. From the common requirements we then define what  
is in scope. Once we know what is in scope we can step by step resolve  
certain characteristics and relations.

I apologise in case the suggested procedure was unclear to anyone.

Tracker, this also should matter to ISSUE-19 and ISSUE-20 ...

Cheers,
	Michael

[1] http://www.w3.org/2011/gld/meeting/2012-02-02#People_UC
--
Dr. Michael Hausenblas, Research Fellow
LiDRC - Linked Data Research Centre
DERI - Digital Enterprise Research Institute
NUIG - National University of Ireland, Galway
Ireland, Europe
Tel. +353 91 495730
http://linkeddata.deri.ie/
http://sw-app.org/about.html

On 3 Feb 2012, at 17:30, Phil Archer wrote:

> Are you sure this is in scope, Michael? You can describe a person in  
> two distinct ways:
>
> - as the individual human being;
> - as an identity.
>
> Individual people have properties that stick with them irrespective  
> of their current activity or location.
>
> Identities change through time, space and context. One can imagine  
> all manner of relationships:
>
> worksFor
> marriedTo
> owns
> partOwns
> onceBakedACakeFor
>
> each one adds to the identity of the person but is very context- 
> specific. In our context, the organisation ontology already has  
> properties for which foaf:agent is the range and I suggest that this  
> is the right approach. Imagine that one day we are tasked with  
> creating a vocabulary for housing. For that we might define things  
> like:
>
> hasMortgageOn
> ownsOutright
> rents
> landlord
> squats
>
> etc. i.e. these all relate something to one or more people and any  
> other domain of interest would have a different set. A cover-all set  
> might be so generic as to be meaningless (knows, is related to,  
> owns, is responsible for?)
>
> To my mind we shouldn't try and do this but just focus on the  
> individual, context-free terms and not the relationships.
>
> It is arguable whether a person's address is something that helps  
> describe the individual or their identity - I'm open to persuasion  
> either way but it's one reason why the ISA Programme has a task  
> force on location/address as well as Person.
>
> Phil.
>
>
>
> On 28/01/2012 10:26, Government Linked Data Working Group Issue  
> Tracker wrote:
>>
>> ISSUE-18: What are the target entities we want relate people to?  
>> [People]
>>
>> http://www.w3.org/2011/gld/track/issues/18
>>
>> Raised by: Michael Hausenblas
>> On product: People
>>
>> We need to define the entities we want to relate people to, such as  
>> to other people via FOAF, online posts using SIOC, etc.
>>
>>
>>
>>
>
>
> -- 
>
>
> Phil Archer
> W3C eGovernment
> http://www.w3.org/egov/
>
> http://philarcher.org
> +44 (0)7887 767755
> @philarcher1
>

Received on Friday, 3 February 2012 18:38:10 UTC