W3C home > Mailing lists > Public > public-gld-wg@w3.org > May 2012

Re: Implementation questions about the ORG Ontology

From: Dave Reynolds <dave.e.reynolds@gmail.com>
Date: Wed, 23 May 2012 09:45:31 +0100
Message-ID: <4FBCA3AB.5080605@gmail.com>
To: Bart van Leeuwen <Bart_van_Leeuwen@netage.nl>
CC: W3C GLD WG WG <public-gld-wg@w3.org>
Hi Bart,

Partial reply for now. When I get some time free to look at it properly 
I'll hope to offer a better response.

The basic use case of describing an organizational structure without 
having to have named individuals is definitely a reasonable one.

I know that the UK use of org for "organograms" did achieve this by 
defining the notion of a "post"[1]. In their case, if I recall 
correctly, a post was in fact treated as an organization because there 
are cases where a post is in fact fulfilled by multiple people or indeed 
a committee. So there when someone takes up a post they have a 
membership relationship with the post as well as to organization 
containing the post.

It might be that some extension like this should be added to org, or at 
least pointed out as a design pattern.

[The alternative of having reporting relations between roles may be 
simpler but has some problems and wouldn't cope with the complexities of 
UK government.]

When I get time I'll dig out the UK extensions and circulate some 
information on them and any further thoughts.

Cheers,
Dave

[1] Those extensions were developed by Jeni Tennision.

On 22/05/12 09:33, Bart van Leeuwen wrote:
> Hi GLD-WG.
>
> As indicated in my earlier mail I'm now looking at implementation
> scenarios for the various products of our working group.
>
> One of the things we would like to do is to have a structure to exchange
> information about 'Chain of command' situations.
> In my work, organizations are not build up around individuals, but
> around roles people have. This counts for our own organization, as much
> as I love my job I'm not there 24/7, but also for the organizations we
> deal with.
> We are interested in exchanging details about the structure of a
> organization without directly filling out the names.
>
> To give two small examples.
> 1) Hotels
> Hotels need to have a evacuation staff, in case of fire alarms they have
> several tasks in conjunction with the fire department, the size of this
> group might vary over the time of day.
> We are not so interested in all the names of people who are "fire
> department contacts" in crisis situation, because only one of them will
> be on the scene anyway, or will play that role during an incident.
> The same applies to the evacuation staff, we are not interested in the
> individuals, but more the numbers, if during a night shift only 2 people
> with the 'evacuation' role are present we need to scale up on our side.
> A schema of which roles are there on specific times of the day and how
> their internal reporting structure is organized would be interesting
> information.
>
> 2) Cross border incidents
> If a incident happens on the border between two countries the parties on
> both sides would like to know what the chain of command is on the other
> side, which person has the credentials to make political / public
> decisions, who should we talk to, and what is the status of their role
> against the roles on this side of the border. Since roles are identified
> through SKOS:Concepts its really easy to formalize the comparison of the
> roles. A Crisis staff in general is assembled from people who are
> available or on duty at that time, the formation of a crisis staff puts
> a individual in the position of a decision making role. In practice this
> means that on forehand its now known who the exact individual will be
> during a time of crisis, we just know this role needs to be fulfilled
>
>  From what I understand from the ORG ontology documentation is that I
> can draw this organizational schema by tracking down people and then
> draw the lines from Role to Role.
> However there doesn't seem to be a way to draw a organizational
> structure without names, but with a command chain. What does sort of
> work is to have anonymous foaf:Agents but that doesn't feel right since
> a foaf:Agent actually represents a instance which is not there yet. Also
> there are situations where the Role is present in the schema, but at the
> specific time there is no one fulfilling it.
>
> Before making comments on changing the ontology itself I wanted to be
> sure I didn't miss something in the documentation, completely
> misunderstand the point or that this use case is out of scope.
>
>
> Met Vriendelijke Groet / With Kind Regards
> Bart van Leeuwen
> @semanticfire
>
> ##############################################################
> # netage.nl
> # http://netage.nl <http://netage.nl/>
> # Enschedepad 76
> # 1324 GJ Almere
> # The Netherlands
> # tel. +31(0)36-5347479
> ##############################################################
Received on Wednesday, 23 May 2012 08:46:28 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 25 June 2013 15:04:57 UTC