Re: [DUV] A sample in Turtle to understand the model from data.gouv.fr

Eric, Ghislain,

*# Q2: Could we add the number of downloads? Maybe as a property of
duv:DatasetUsage? or maybe in duv:Rating?*

I think feedback, in a broader sense, could embraces all the data usage.
But I agree with you, Eric, in restricting the term feedback in DUV to the
human interactions, the comments and discussions that humans will post.
Maybe we can use a term like statistics or measures to these numeric
things, as for example, duv:UsageStatistics or duv:UsageMeasures. VoID uses
the term statistics to define properties about numbers of triples, classes,
etc. of a Dataset (http://www.w3.org/TR/void/#statistics). Maybe we can use
DAQ concepts of measures and observations.

Best,
Laufer

2015-04-29 6:50 GMT-03:00 Eric Stephan <ericphb@gmail.com>:

> Ghislain,
>
> Fantastic!  I will send you a link to our DUV wiki page which really needs
> updating.
>
> My comments from iPhone below...
> Eric S
>
> Sent from my iPhone
>
> On Apr 29, 2015, at 2:01 AM, Ghislain Atemezing <
> auguste.atemezing@eurecom.fr> wrote:
>
> Hi all,
>
> I’ve tried to use the current model of DUV to capture a small use
> (compared to DBpedia one) taken from data.gouv.fr.
> In this French portal, providers can see the usage of their datasets from
> the users with account, where they can provide applications, blogs or any
> reuse of the original datasets.
> I took the case of this dataset [1] in heath domain.
> The first sample is available here [2] (*) and from this exercise I have
> some questions below regarding the model:
>
> *# Q0: Can we also add the dcat:Catalog as in this use case?*
>
> +1 I believe we discussed this at our F2F meeting this seems very
> appropriate.
>
> *# Q1: Is it worth to add a dcat:Distribution as another duv:Feedback?*
>
> Interesting...this might fit into examples like Yaso mentioned to support
> curation activities?
>
> *# Q2: Could we add the number of downloads? Maybe as a property of
> duv:DatasetUsage? or maybe in duv:Rating?*
>
> There are a number of metrics like this we have talked about , up until
> now it has been associated as a type of feedback?
>
> I'm not entirely opposed to thinking about this but I think we need to
> make clear lines of distinction between what we mean when we say usage
> versus feedback.  If we can put metrics in usage and give a clear rationale
> then this may be ok.
>
> *# Q3: In which case can we use the duv:annotes(foaf:Agent, dcat:Dataset)
> property ? *
>
> We haven't entirely evolved this idea other than a general reify notion of
> who provides the annotation. This needs to be flushed out more.
>
>
> IMO, having such a sample can help us understanding better what we
> can/need further in our model.
>
> HTH
>
> +1 HUGE THANK YOU! :-)
>
>
> Ghislain
> (*) Probably I should put in somewhere in our Wiki page?
>
> [1]
> https://www.data.gouv.fr/fr/datasets/depenses-d-assurance-maladie-hors-prestations-hospitalieres-donnees-nationales/
>
> [2] https://gist.github.com/gatemezing/1349ce643d90fa49cdca
>
>
> ---------------------------------------
> Ghislain A. Atemezing, Ph.D
> ghislain.atemezing@gmail.com
> Google+: http://google.com/+GhislainATEMEZING
> Twitter: @gatemezing
>
>
>
>
>
>


-- 
.  .  .  .. .  .
.        .   . ..
.     ..       .