Re: Question regarding the Organization Ontology

A position (the thing which can be vaccant) can theoretically be the 
author of a work, but it's not normal.

However a person-in-a-position can be a creator of a work and it can be 
very important to note the "hat" with which the creation was done. As 
linked data gets more linky it's important to keep modelling some 
distinctions. In most cases, my (theoretical, for now) My Little Pony 
fan fiction should not be placed in the same authorship list as my 
papers on open data, even thought the same individual authored them, it 
wasn't with the same hat on.

Of course, now I have a nagging urge to write the tale of "Linky-Pie the 
Open Data Pony" just to illustrate the issue.

Basically a person-in-a-job is a foaf:Agent and can be given a 
meaningfully distinct identifier to the foaf:Person. Many systems would 
just collapse these, but there are bibliographic cases were it's useful 
not to. But only model what you need. For most systems this is going 
doing the path to modelling things in such a perfect way that nobody 
will ever use your data.


On 10/03/2014 21:20, Chris Beer wrote:
>
> How so? A report can be created by a person holding a role, but the 
> role cannot create a report if it is vacant... It is just a thing in 
> this context and cannot be an agent.
>
> Happy to be shown otherwise?
>
> Cheers
>
> Chris Beer
> Australia
>
> Sent from my Sony Xperia™ smartphone
>
>
>
> ---- Christopher Gutteridge wrote ----
>
> There's no reason the creator can't be both the individual *and* the 
> role. Both are true. Both are useful.
>
> In EPrints, we decided to provide both direct and easy to consume 
> dcterms:creators for authors, but also a bibo:authorList which links 
> to an RDF sequence, for people who need to know the ordering of the 
> authors. Both are valid and it's fine to include both in one scheme.
> http://eprints.soton.ac.uk/cgi/export/eprint/191579/RDFN3/eps-eprint-191579.n3
>
> Your situation is similar. If you really can't decide, just do both. 
> Although I'd strongly recommend writing a simple application to 
> consume your own data. It's amazing how quickly things come into focus 
> when you actually try to use your beautifully modelled graph and 
> discover it's missing vital bits or way to complex.
>
> On 10/03/2014 18:07, James McKinney wrote:
>> Hmm, I guess that will depend on your model for the report. I suppose 
>> one way of doing it would be to have the range of the "creator" 
>> property be either an individual or an org:Membership? Another way 
>> may be to leave the "creator" property as-is, and to create a new 
>> property on the report to point to the relevant membership (or more, 
>> generally, the context in which the creator created the report); that 
>> may be the better option as it won't complicate the creator property. 
>> Those are the two ideas that come to mind.
>>
>> James
>>
>> On 2014-03-10, at 1:48 PM, Anne Ward wrote:
>>
>>> Hi James,
>>>
>>> I have one more question regarding this approach. This works 
>>> perfectly for identifying the posts held by an individual and their 
>>> timing.
>>>
>>> My next question relates to a situation where an individual is the 
>>> “creator” of a report (for example) while holding a given post - 
>>> i.e., they can be a “creator” as both an individual (as in private 
>>> letters) and as an individual “holding” a “post” with an 
>>> organization (e.g., letters written while in office).
>>>
>>> How would one model the second situation?
>>>
>>> Thanks in advance.
>>>
>>> Anne
>>> On Mar 10, 2014, at 11:56 AM, James McKinney <james@opennorth.ca 
>>> <mailto:james@opennorth.ca>> wrote:
>>>
>>>> I've done something similar to Bernard, except instead of using a 
>>>> new Position class, I simply add a "post" property to 
>>>> org:Membership. Instead of people holding posts directly, all 
>>>> people hold posts through their memberships; I therefore do not use 
>>>> the org:holds or org:heldBy properties. org:Membership already has 
>>>> org:memberDuring to express the time interval during which the 
>>>> membership exists, or in my case during which the post is held.
>>>>
>>>> With respect to Christopher's affiliations, I add an "onBehalfOf" 
>>>> property to org:Membership, to express on whose behalf that person 
>>>> is a member.
>>>>
>>>> I prefer to avoid a proliferation of sub-classes when an additional 
>>>> property would do.
>>>>
>>>> James
>>>>
>>>> On 2014-03-10, at 11:21 AM, Anne Ward wrote:
>>>>
>>>>> Thank you everyone for your quick responses as well as the options 
>>>>> you have identified.
>>>>>
>>>>> I now will take a closer look at the modelling solutions 
>>>>> identified to better understand them and to identify how each 
>>>>> would address the example I am trying to work out.
>>>>>
>>>>> Anne
>>>>> On Mar 10, 2014, at 10:39 AM, Christopher Gutteridge 
>>>>> <cjg@ecs.soton.ac.uk <mailto:cjg@ecs.soton.ac.uk>> wrote:
>>>>>
>>>>>> It is important to be able to talk about a post which is vacant.
>>>>>>
>>>>>> You potentially need to be able to talk about an individual, a 
>>>>>> post within the organisation and their membership of that post. 
>>>>>> This becomes very useful when you want to distinguish 
>>>>>> relationships and responsibilities.
>>>>>>
>>>>>> For example; "Post 120" supervises "Post 121". However things 
>>>>>> like committee memberships are actually attached to your 
>>>>>> membership of the orgainsation... if person X is on a committee 
>>>>>> and then retires and person Y is appointed to her post, it does 
>>>>>> not automatically make Y a member of the committee, other 
>>>>>> committee memberships may be explicitly for people with certain 
>>>>>> posts.
>>>>>>
>>>>>> Finally you have relationships to or between individuals 
>>>>>> themselves, however these will generally be out of the scope 
>>>>>> about what an organisation cares about.
>>>>>>
>>>>>> One area this has mattered for me is in producing linked data 
>>>>>> from a conference. http://programme.ecs.soton.ac.uk/ -- I ended 
>>>>>> up inventing an "Affiliation" class as I needed to represent the 
>>>>>> same person speaking in two different sessions and with a 
>>>>>> different affiliation. One talk was about his work, the second 
>>>>>> about a hobby project or somesuch. It mattered to represent which 
>>>>>> "hat" he was wearing.
>>>>>>
>>>>>>
>>>>>> On 10/03/14 13:29, jean delahousse wrote:
>>>>>>> Hello,
>>>>>>> Why not use Membership which is richer than Post ?
>>>>>>> http://www.w3.org/TR/2014/REC-vocab-org-20140116/#class-membership
>>>>>>> It is the class I proposed for EU directory.
>>>>>>> Jean
>>>>>>>
>>>>>>>
>>>>>>> 2014-03-10 14:24 GMT+01:00 Bernard Vatant 
>>>>>>> <bernard.vatant@mondeca.com <mailto:bernard.vatant@mondeca.com>>:
>>>>>>>
>>>>>>>     Hi Anne
>>>>>>>
>>>>>>>     What I do for that kind of situation is to make distinct
>>>>>>>     classes "Position" and "Post" (or Job, whatever you want to
>>>>>>>     name it)
>>>>>>>
>>>>>>>     :Anne :positionHeld :Position12345
>>>>>>>     :Position12345  :beginDate "2012-10-01"
>>>>>>>     :Position12345  :endDate "2013-12-31"
>>>>>>>     :Position12345 :postHeld  :PostX
>>>>>>>     :Position12345  :employer :OrgY
>>>>>>>
>>>>>>>     :Position12345 is actually an "Event"
>>>>>>>     :PostX is qualifying the "Position type" or "Job", e.g.;
>>>>>>>     "Chief Technical Officer" "Documentalist" etc.
>>>>>>>
>>>>>>>     You can relate successive positions held by the same person
>>>>>>>     using something like http://vocab.org/bio/0.1/.html
>>>>>>>
>>>>>>>     My 0.02
>>>>>>>
>>>>>>>     Bernard
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>     2014-03-07 17:14 GMT+01:00 Anne Ward <anne.ward@rogers.com
>>>>>>>     <mailto:anne.ward@rogers.com>>:
>>>>>>>
>>>>>>>         Hi,
>>>>>>>
>>>>>>>         I am planning to use the organization ontology in
>>>>>>>         examples of defining relationships between persons and
>>>>>>>         organizations. In particular, I found the addition of
>>>>>>>         “Post” quite applicable to the examples I am trying to
>>>>>>>         illustrate.
>>>>>>>
>>>>>>>         I have a question regarding its usage, when specifying
>>>>>>>         that a person “holds” a “Post” within an organization.
>>>>>>>         As a “Post” can be held by many people over time, what
>>>>>>>         would be the best approach for modelling the time
>>>>>>>         interval in a which a given person “holds” a given “Post”?
>>>>>>>
>>>>>>>         Please advise.
>>>>>>>
>>>>>>>         Thank you.
>>>>>>>
>>>>>>>         Anne Ward
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>     -- 
>>>>>>>     *Bernard Vatant
>>>>>>>     *
>>>>>>>     Vocabularies & Data Engineering
>>>>>>>     Tel : + 33 (0)9 71 48 84 59
>>>>>>>     Skype : bernard.vatant
>>>>>>>     http://google.com/+BernardVatant
>>>>>>>     --------------------------------------------------------
>>>>>>>     *Mondeca*****
>>>>>>>     3 cité Nollez 75018 Paris, France
>>>>>>>     www.mondeca.com <http://www.mondeca.com/>
>>>>>>>     Follow us on Twitter : @mondecanews
>>>>>>>     <http://twitter.com/#%21/mondecanews>
>>>>>>>     ----------------------------------------------------------
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> -- 
>>>>>>> Jean Delahousse
>>>>>>>
>>>>>>> ----------------------------------------------------------------------------------------------------
>>>>>>> delahousse.jean@gmail.com <mailto:delahousse.jean@gmail.com> - 
>>>>>>> +33 6 01 22 48 55
>>>>>>> http://fr.linkedin.com/in/jeandelahousse
>>>>>>>
>>>>>>>
>>>>>>
>>>>>> -- 
>>>>>> Christopher Gutteridge --http://users.ecs.soton.ac.uk/cjg
>>>>>>
>>>>>> University of Southampton Open Data Service:http://data.southampton.ac.uk/
>>>>>> You should read the ECS Web Team blog:http://blogs.ecs.soton.ac.uk/webteam/
>>>>>>
>>>>>> Would you recommend the software you use to another institution?
>>>>>> http://uni-software.ideascale.com/
>>>>>
>>>>
>>>
>>
>
> -- 
> Christopher Gutteridge --http://users.ecs.soton.ac.uk/cjg
>
> University of Southampton Open Data Service:http://data.southampton.ac.uk/
> You should read the ECS Web Team blog:http://blogs.ecs.soton.ac.uk/webteam/
>

-- 
Christopher Gutteridge -- http://users.ecs.soton.ac.uk/cjg

University of Southampton Open Data Service: http://data.southampton.ac.uk/
You should read the ECS Web Team blog: http://blogs.ecs.soton.ac.uk/webteam/