Re: DWBP vocabularies and annotations

Hi Eric,
First, let me say that I realized today I have been reading the graph wrong. I didn't realize that the closed arrows were indicating a subclass. So, I think we're actually on the same page regarding feedback. But I do think that other types of usage that people would want to report make sense as annotations as well. I guess I'm thinking of annotations as a generally useful construct for this because they form a type of formal reference back to the original that doesn't exist elsewhere on the web. Since I believe a major goal of the vocabulary is to enable users of data to report that usage back to publishers in a way that the publishers can find, I see annotations as a mechanism for allowing that sort of reference. I think publishers would be interested in being made aware of all the types of use we're considering, so that makes me think perhaps they should all be defined as some type of annotation. I also like that annotations already have a way to record provenance.
What do you think?
-Annette

On May 29, 2015, at 8:17 AM, Eric Stephan <ericphb@gmail.com> wrote:

> Annette,
> 
> First of all thank you for pointing us to the Web Annotations Motivations in April after our F2F.  I have found this to be very useful.  Our approach has been to make a subclass of duv:Feedback http://w3c.github.io/dwbp/vocab-du.html#vocabulary-specification so that we inherit all aspects of the Annotation class.  
> 
> I am reluctant to just use the Annotation class as a substitute for our subclass solution, because we are trying to model specific requirements for feedback on datasets as opposed to a generic annotation about any web resource.  
> 
> Without a subclass it becomes clumsy from a search and discovery standpoint separating feedback about the dataset from generic annotations, because I would lack a crucial part of a query pattern needed to hone in on what I'm after and any of my queries would require a secondary step for filtering any annotation that was not related directly to the aspect of feedback. 
> 
> Another solution just using generic annotations is annotating the annotation to declare that this feedback on data usage, but again declaring a subclass does that for me.  
> 
> Does this make sense to you?
> 
> Cheers,
> 
> Eric
> 
> On Thu, May 28, 2015 at 11:13 PM, Annette Greiner <amgreiner@lbl.gov> wrote:
> All the talk about defining extensions to DCAT is making me re-examine my thoughts about how the DUV relates to annotations. I keep thinking that we should be defining our usage annotations as oa:Annotations in the web annotations vocabulary, rather than duv:UsageAnnotations. I think we and the annotations WG are both trying to define a means of linking additional content to existing web resources, in our case datasets. The annotations spec defines a list of motivations[1], some of which I think already represent concepts that we're talking about including, "commenting", "describing", "editing", and "questioning", for example.  I think we could define new motivations for what's missing, like "citing" and "rating". We really should avoid creating a situation where people need to choose between one annotation system and the other, IMHO.
> -Annette
> 
> [1] http://www.w3.org/TR/2014/WD-annotation-model-20141211/#motivations
>

Received on Friday, 29 May 2015 16:26:51 UTC