Re: Status on Action-174 ?

Annette,

Great thoughts here, I've provided comments, see what you think :-)

Eric

On Thu, Apr 23, 2015 at 12:54 PM, Annette Greiner <amgreiner@lbl.gov> wrote:

> It's good to see some ideas getting put out there for these terms. I do
> have a couple of concerns with the definitions here. First, I think we need
> to avoid definitions that use the same term they are defining. The
> definition of "consumer" here is tautological, relying on the words
> "consumer" and "consuming".
>
> -- Good points, yes we must be vigilant!   We need to make use of the
glossary big time.


> Second, I think the term "citation" suggests something narrower. In
> academia, if Alice writes a paper and Bob cites it, Bob is saying to the
> world that he used Alice's work. If Alice cited her own work in the work
> itself, it would be pointless. "Citation" doesn't cover the case of the
> originator of the work stating that someone else used it.
>
> --  I think the other case for citation is providing a link describing how
you want others to cite it.

An awful lot of what we've been talking about is already defined in the
> annotations model [1]. For feedback, the current web annotations draft
> already covers that pretty well. Take a look at their motivations [2]. We
> might suggest additional motivations, but there are already "commenting",
> "editing", and "questioning".
>
> -- While the Annotation model does cover it in a very general way thus
giving rise to the concern that there might be large interpretations of how
I think of feedback solely relying on Annotations, I am attracted to the
SIOC feedback model because it was built specifically to represent feedback
in forums. By selecting a common model for feedback, I argue that an
explicitly declared vocabulary greatly increases the chances of making
dataset feedback more discoverable because consumers can correlate and
cross reference feedback from different dataset forums using a consistent
query pattern.  The Annotation model is so general that cross referencing
forums represented in a variety of ways would make discovery of feedback
more difficult.


> I think that what we should be focusing on is usage annotations, possibly
> just an additional motivation for the annotations model. Usage annotations
> can be the same for the publisher and the re-user. In both cases, someone
> is creating an annotation outside the original work (outside the published
> dataset) that says that particular dataset was reused by someone else. So,
> if Alice publishes a dataset and Bob uses it in creating a visualization,
> which Carol views online, Bob needs a way to tell the world that he used
> Alice's data, so that Carol can see it and so that Alice can be made aware
> of it. Alice could make the same annotation as Bob in her role as a
> webizen; she needn't adopt her role as the original publisher to do so.
> There is no need to distinguish between publisher and consumer. Either way,
> Carol can see it and Alice is also aware of it.
>
> -- I agree that we should also use Usage annotations in the way that you
described.  I agree about the blur between publisher and consumer in
situations like these.


> -Annette
>
> [1] http://www.w3.org/TR/annotation-model/
> [2] http://www.w3.org/TR/annotation-model/#motivations
>
> On Apr 23, 2015, at 10:01 AM, Ig Ibert Bittencourt <ig.ibert@gmail.com>
> wrote:
>
> > Hi Yaso and Eric,
> >
> > Wonderful you've started the glossary. I would suggest some updates in
> the definition of the following concepts:
> >       • Data Producer: this entity represents an agent (human or
> machine) in the role of a data producer responsible for producing data (or
> even the dataset) in any phase of the data life cycle. It is important to
> say the production of new data can happen through the consumption/reuse of
> old data;
> >       • Data Consumer: this entity represents an agent (human or
> machine) in the role of a data consumer responsible for consuming data from
> one or more datasets.
> >       • Data Publisher: this entity represents an agent (human or
> machine) in the role of data publisher responsible for publishing one or
> more datases. It is important to say that this role is different from data
> producer, although they can be owned by the same agent;
> >       • Citation: Citations is an action performed by a citing entity to
> a cited entity, qualified/characterized as direct and explicit, indirect or
> implicit [1]. In the context of DWBP, it is a formal feedback
> (bibliographic reference in the form of published material as a book,
> paper, web page) performed by an agent (in the role of a data consumer or
> data publisher) to a dataset.
> > Please, let me know if you agree of not.
> >
> > [1] http://purl.org/spar/cito
> >
> > Cheers,
> > Ig
> >
> > 2015-04-23 12:33 GMT-03:00 <yaso@nic.br>:
> > Hi Eric!
> >
> > I'm glad you asked. I made a very rough draft and uploaded to my fork on
> github because the doc is still with no styling and proper IDs. I sent it
> to Phil and the editors to get some feedback, but had no answer till now.
> >
> > Basically, I picked the idea of that conversation at the f2f and
> threshed the details of the Data on the Web Life cycle proposed by
> Bernadette. The draft is at [1] and you can see it rendered at [2].
> >
> > What I am proposing is that we use actions that transforms data in to
> something else, like a dataset or metadata, as turning points to divide
> whether someone is a publisher or a consumer. Furthermore, as a data
> creator can be also someone that collects data from others using software,
> like facebook or yandex, I propose that we focus on 3 "representations" of
> data: data (as raw data), dataset (as encoded file or structured dataset)
> or metadata (whatever is the format).
> >
> > Given this way of thinking the cycle of data on the web, data archiving
> techniques and data preparation or data planning are out of the scope of
> this WG, thought 303 pages and 404 are in. Data encoded in file formats is
> out of the scope also, but only if people involved with data mining and
> enriching wants.
> >
> > The idea still have to be polished, but I think it is a good way out for
> our abstractsss discussions to focus on actions performed to delineate
> concrete lines of definitions.
> >
> > I'm keen for the feedback of the WG.
> >
> > [1] https://github.com/yaso/dwbp
> > [2] http://yaso.is/dwbp/glossary.html
> >
> >
> > Cheers
> > Yaso
> >
> >
> >
> >
> >
> > Quoting "Eric Stephan" <ericphb@gmail.com>:
> >
> > Hi Ig,
> >
> > There are quite a bit of definitions floating around, I was wondering if
> > you needed help on this task.  Please let me know.
> >
> > Eric S
> >
> >   https://www.w3.org/2013/dwbp/track/actions/174
> >
> >
> >
> >
> >
> >
> >
> > --
> >
> > Ig Ibert Bittencourt
> > Professor Adjunto III - Instituto de Computação/Universidade Federal de
> Alagoas (UFAL)
> > Vice-Coordenador da Comissão Especial de Informática na Educação
> > Líder do Centro de Excelência em Tecnologias Sociais
> > Co-fundador da Startup MeuTutor Soluções Educacionais LTDA.
>
>