Re: dwbp-ISSUE-154 (Eric Stephan): Handling data usage references to collection of datasets [Data Usage Vocabulary]

Just to document what other efforts under the data activity are doing, I
wanted to document that the CSV on the Web Working Group Tabular Data Model
[1] has a requirement to associate metadata with a group of tables (this
could be done within a dataset), and a directory of datasets.  While this
is a very specific implementation it does illustrate the need outside the
scientific community to associate metadata to a group of datasets.

Thanks,

Eric S

References
[1] http://www.w3.org/TR/tabular-data-model/#locating-metadata


On Fri, Apr 10, 2015 at 9:26 AM, Data on the Web Best Practices Working
Group Issue Tracker <sysbot+tracker@w3.org> wrote:

> dwbp-ISSUE-154 (Eric Stephan): Handling  data usage references to
> collection of datasets [Data Usage Vocabulary]
>
> http://www.w3.org/2013/dwbp/track/issues/154
>
> Raised by: Eric Stephan
> On product: Data Usage Vocabulary
>
> Dataset is currently defined along the lines of the DCAT definition.  It
> is fairly common that organizations or scientific domains refer to a
> dataset as a collection of data.  How do we accommodate the need to provide
> feedback on a collection of data?
>
> For instance if I am performing a simulation of climate change and sharing
> the results with a collaborator they will be mainly providing feedback on
> the entire collection of datasets and not each dataset individually.
>
> If we use dcat:Dataset and dcat:Catalog without breaking the DCAT model
> then we basically have to duplicate all properties in the data model.  I'd
> like to avoid this.
>
> What I propose is that we have:
> 1) a duv:Collection class as you suggested Bernadette
> 2) make a parent class for dcat:Dataset and duv:Collection that would
> serve as a "inode"
> 3) Call this parent class duv:Data (or something else - please suggest)
> 4) All data usage and feedback would relate to duv:Data.
>
> Also we could consider:  schema.org  tends to use a parent class called
> CreativeWork that carries many of the aspects of data usage.
> CreativeWork (https://schema.org/CreativeWork ) acts as a parent class of
> anything on the web including DCAT.  It also provides mechanisms for
> citation, discussion, etc. http://schema.org/docs/schema_org_rdfa.htm
>
>
>
>
>
>