Re: Some comments on Actions

Hi Leigh,

+1 for your remarks re. the new schema.org Actions proposal and your 
recommendation of the Event Ontology. Although, I think that the new 
actions proposal goes even a bit further into a hypermedia driven 
environment (which is really interesting and well-though-out in my mind).

Generally, we can also make aware of the siblings or synonyms of Action 
a.k.a. Event, Activity, Affordance, Intent (maybe there are even more).

Besides, I would like to recommend a further ontology in this context 
that I've co-designed some time ago - the Counter Ontology [1]. This 
ontology contains a class called co:ScrobbleEvent [2] that somehow bears 
the same meaning what I think the Actions proposal would like to address 
as well, or? There are also examples how this class can be utilised in a 
more specific context, such as play back and skip counter [3], and how 
it can be extended as well, such as pbo:SkipEvent [4, 5] ;)

Finally, re. the discussion of "object" and "target" for Actions, I 
would like to suggest the following terms (based on Jim's example):

AuthenticatedUser (actor/subject) can post ((tentative) action/verb) 
FormX  (object) to SystemY (context)

... and like Leigh remarked, the result (product) is also often 
important. One should keep in mind that not every attribute needs to 
mandatory, e.g., object and context could be optional.

Cheers,


Bo


[1] http://purl.org/ontology/co/core#
[2] http://purl.org/ontology/co/core#ScrobbleEvent
[3] 
http://purl.org/ontology/pbo/playbackontology.html#sec-play-back-and-skip-counter-example
[4] http://purl.org/ontology/pbo/core#SkipEvent
[5] 
http://purl.org/ontology/pbo/playbackontology.html#sec-extended-skip-counter-example

On 5/17/2013 6:51 PM, Leigh Dodds wrote:
> Hi,
>
> Some comments/thoughts on the Actions draft.
>
> Reading the draft I wondered whether there ought to be clearer
> separation between:
>
> * the Action
> * the object to which the action applies
> * the result of performing that action
>
> That might be the actual goal, but I found some of the examples
> confusing. For example BuyAction seems to be a type of action, but the
> document also shows things that are of type BuyAction but which are
> actually the *result*.
>
> This is also in evidence in examples like:
>
> Thing > Action > BuyAction > BuyTicketAction > BuyMovieTicketAction.
>
> If there's a clean separation between the different objects then you
> wouldn't need to have extensions for Tickets or Movies: these are the
> objects that are being bought. An individual service (or app) will
> relate the generic BuyAction to one or more types of object, e.g. a
> "buy ticket action" in an instance of a BuyAction that has been
> described as applying to Ticket objects.
>
> For Action results it might be useful to look at the Event Ontology
> [1]. That has a good general model for describing an event, its
> participants, the things involved (the Application or Service here),
> and the outputs, e.g. a Reservation.
>
> Schema.org could use a similar base model which could then be
> extended. Again the types would be fairly generic. E.g. an
> Order/Purchase could apply to a Ticket, Reservation, Product, etc. It
> would be a common basis for extension by others.
>
> Cheers,
>
> L.
>
> [1]. http://motools.sourceforge.net/event/event.html
>
>
> --
> Leigh Dodds
> Freelance Technologist
> Open Data, Linked Data Geek
> t: @ldodds
> w: ldodds.com
> e: leigh@ldodds.com
>

Received on Friday, 17 May 2013 23:31:42 UTC