This is the archive of the goodrelations dicussion list

Hi Osma,

Interesting modeling scenario.
I see the motivation to the changes you propose to gr:Location,
however as you mention, gr:Location is intended to represent a real,
physical location. In that sense, under the current definition,
regarding question:

1. Subclassing gr:Location to represent non-physical locations can be
seen as contradictory because "a non-physical location" is not really
a subtype/subclass of a "physical location".

2. I believe changes to the definition of gr:Location (to represent
both physical and non-physical locations) are not being considered for
now, afaik...

However, there is an approach that could address your modeling
scenario and that would fit into the current GRs schema.
You could extend the class gr:DeliveryMethod with new individuals (or
even classes, depending on your ultimate needs) to represent the means
of providing (offering) the product or service that you mention (day
care, medical assistance, library), i.e. something along the line of
foo:DeliveryMethodPhone, or foo:DeliveryMethodWeb, etc.

You could also add to that, an additional extension similar to the
approach used in the Accommodation Ontology (a curated extension to
GRs in that domain -- see: http://purl.org/acco/ns#) with the object
property acco:availabilityTimes that relates a specific accommodation
feature with the particular gr:OpeningHoursSpecification (some
examples here: http://ontologies.sti-innsbruck.at/acco/ns.html#Examples).

It would take some more time and a bit more information about the
modeling scenarios, to bring together the conceptual elements of these
extensions to the GRs ontology to craft examples in the context of the
3 modeling cases you mention, (day care, medical assistance, library),
however in the meantime, I thought I would point out these two ideas
in case they help.

Best,
Bene Rodriguez
-- 
Research Associate
E-Business and Web Science Research Group
Department of General Management and E-Business
Universitaet der Bundeswehr Munich (Germany)
phone: +49 89 6004-2849
email: benedicto.rodriguez at ebusiness-unibw.org
web:   http://purl.org/beroca

On Tue, Nov 27, 2012 at 11:34 AM, Osma Suominen <osma.suominen at aalto.fi> wrote:
> Hello,
>
> I'm developing a vocabulary/ontology for describing public services. The
> goal is to express things like:
>
>  * this kindergarten at this location provides day care services for
> families living within this municipality; its opening hours are X, location
> Y and contact information Z
>  * the cities A,B,C together offer this free help line that residents can
> call to get medical assistance; the service is actually provided by medical
> company X (paid for by the cities)
>  * the city library in X provides library services at this location; on its
> website at URL Y, you can also borrow e-books as a self-service
>
> As you might guess, many parts of the ontology are heavily influenced by,
> and derived from, GoodRelations. It is possible to express most of the
> important things in this domain using GR constructs such as gr:Offering,
> gr:BusinessEntity and probably some form of gr:ProductOrService subclasses.
> There will also be extensions detailing the various relationships between
> organizations (e.g. who pays for what) and for accessibility information.
>
> What I'm struggling with is that some of the services are provided in
> non-physical service points/channels, i.e. on the web or by phone. In many
> cases, the "same" service is available through more than one channel.
> Opening hours are relevant for both physical locations and phone services,
> in rare cases even for websites.
>
> I'd like to use gr:availableAtOrFrom and gr:Location to express also these
> virtual service points and then (optionally) attach opening hours
> information to them using gr:hasOpeningHoursSpecification. But part the
> gr:Location description says "Locations are characterized by an address or
> geographical position and a set of opening hour specifications for various
> days of the week." In my public service case, the non-physical service
> points do not have a meaningful geographical address or position. Instead
> the "location" is a URL of a web site or a phone number. I can imagine
> similar use cases in the e-commerce domain that GR is aimed at, for example
> companies that sell downloads (magazine articles, music, creditworthiness
> reports...) on the web.
>
> So finally to my questions:
>
> 1. Do you think it would still be fine to use (a subclass of) gr:Location to
> model these?
>
> 2. If yes, could the description of gr:Location be amended to include also
> non-physical locations offering products and services?
>
> Thanks in advance,
> Osma
>
> --
> Osma Suominen | Osma.Suominen at aalto.fi | +358 40 5255 882
> Aalto University, Department of Media Technology, Semantic Computing
> Research Group
> Room 2541, Otaniementie 17, Espoo, Finland; P.O. Box 15500, FI-00076 Aalto,
> Finland
>
> _______________________________________________
> goodrelations mailing list
> goodrelations at ebusiness-unibw.org
> http://ebusiness-unibw.org/cgi-bin/mailman/listinfo/goodrelations