Re: Using schema.org Dataset metadata properties

On 15 Sep 2014, at 19:57 , Jeni Tennison <jeni@jenitennison.com> wrote:

> Ivan,
> 
> Given that we’re adopting JSON-LD for the metadata file, anyone *can* use any vocabulary. I was thinking that we should including the binding of ‘dc’ to the Dublin Core namespace so that people can easily add metadata in that scheme if they want to.
> 

+1. We will have a JSON-LD @context anyway, so adding this is obviously good. And, to echo Gregg's comment on this, that should be done with 'schema' as well, whatever the outcome of this thread.


> I think there is huge value in having a predictable structure to metadata, as it helps with validation, display and conversion. Adopting JSON-LD in effect enforces a particular structure, eg saying that “publisher” must look like:
> 
>   “publisher”: {
>     “@id”: "http://www.hefce.ac.uk/“,
>     “name": "Higher Education Funding Council for England (HEFCE)"
>   }
> 
> or
> 
>   “publisher”: "http://www.hefce.ac.uk/“
> 
> and not
> 
>   “publisher”: "Higher Education Funding Council for England (HEFCE)"
> 
> Adopting schema.org normatively would mean saying that “publisher” means what it means in schema.org, which I think is what we would want to do.


We can still define the @context file in a way that "publisher" is mapped on "schema.org/publisher". But if we consider that normative, that means any implementation MUST check the right structure for all the terms that we choose to add. Ie, MUST test the correct usage of the terms as defined by schema.org. And it also _suggests_ that authors MUST use schema.org terms for something like a publisher. 

However... there may be problems relying on schema.org that heavily. Making a normative reference to an external document from a W3C Standard means that the referred document or specification has to fulfill some requirements as for the decision process used to define the technology, the stability and versioning, the IPR status, etc. We know that, in other groups, the discussions it generates to refer to WHATWG documents, and the issues around those have escalated beyond the originating Working Group and are discussed by the TAG, the W3C management, etc. This issue has not come up, so far, for schema.org, but by adopting schema.org as a normative reference, we will have to face the same issues.

(Referring to IETF documents, ISO standars, Dublin Core vocabularies, and some others do not represent the same challenges; their governance is very different.)

If the terms are considered to be informative, implementations MAY issue warnings to an incorrect usage of the term if the authors choose to use it. It is not so far off what we need.

Ivan


> 
> Cheers,
> 
> Jeni
> 
> -----Original Message-----
> From: Ivan Herman <ivan@w3.org>
> Reply: Ivan Herman <ivan@w3.org>>
> Date: 14 September 2014 at 08:07:03
> To: Jeni Tennison <jeni@jenitennison.com>>
> Cc: W3C CSV on the Web Working Group <public-csv-wg@w3.org>>
> Subject:  Re: Using schema.org Dataset metadata properties
> 
>> I have a meta-question on this. Is the list of terms listed in the document normative or  
>> informative? The current document does not make a difference (ie, by default, it is normative,  
>> including the references), but I presume this is simply because we never asked ourselves  
>> the question.
>> 
>> At the moment, the text says:
>> 
>> [[[
>> Descriptions may contain any properties defined by [DC-TERMS] to describe the table.  
>> This specification does not define any application behaviour associated with these  
>> properties being present, except that validation of metadata files must check that,  
>> if they are present, they adhere to the syntax defined here.
>> ]]]
>> 
>> This at first suggests that the [Dublin Core] vocabulary is informative (and optional)  
>> but then it mandates specific value syntax for some of the properties when validating.  
>> I think it could be debated whether this additional validation requirement actually  
>> makes the reference normative, but it is not clear. I guess the question is whether we  
>> will have a notion of conforming metadata, of a possible metadata validator, and what  
>> they are supposed to exactly do.
>> 
>> Why is this question relevant? Because if the whole section is normative than we MUST  
>> make a choice on whether, for a specific goal, we choose DCTERM or schema. If it is informative,  
>> there is no problem referring to both and let the end user decide (and, actually, the exact  
>> value syntax issue could also be removed simply referring to the definition of these  
>> terms by DCMI and schema.org, respectively.)
>> 
>> (There is also an editorial/W3C issue. There are fairly stringent rules on whether we  
>> can refer, _normatively_, to an external document. While this is not a problem with DCTERM,  
>> this has not yet done before for schema.org, and it may lead to some discussions...)
>> 
>> Ivan
>> 
>> 
>> 
>> On 13 Sep 2014, at 18:28 , Jeni Tennison wrote:
>> 
>>> Hi,
>>> 
>>> In the current metadata document here:
>>> 
>>> http://w3c.github.io/csvw/metadata/#common-properties
>>> 
>>> the spec maps adopts the list of Dublin Core properties for describing tables etc. As  
>> ISSUE 6 says, this might not be the right choice: there might be other standard vocabularies  
>> that should be used instead or as well.
>>> 
>>> On the call this week, Dan suggested using schema.org instead, namely the properties  
>> on Dataset here:
>>> 
>>> http://schema.org/Dataset
>>> 
>>> The properties there are informed by DCAT which itself was informed by Dublin Core.  
>>> 
>>> Any thoughts?
>>> 
>>> Jeni
>>> 
>>> -----Original Message-----
>>> From: CSV on the Web Working Group Issue Tracker  
>>> Reply: CSV on the Web Working Group >
>>> Date: 10 September 2014 at 13:23:37
>>> To: jeni@jenitennison.com >
>>> Subject: ACTION-26: Write to mailing list re using schema.org rather than dublin core  
>> for metadata about csv files, then binding decision on following telcon (CSV on the Web  
>> Working Group)
>>> 
>>>> ACTION-26: Write to mailing list re using schema.org rather than dublin core for metadata  
>>>> about csv files, then binding decision on following telcon (CSV on the Web Working  
>> Group)
>>>> 
>>>> http://www.w3.org/2013/csvw/track/actions/26
>>>> 
>>>> On: Jeni Tennison
>>>> Due: 2014-09-17
>>>> 
>>>> If you do not want to be notified on new action items for this group, please update your  
>>>> settings at:
>>>> http://www.w3.org/2013/csvw/track/users/33715#settings
>>>> 
>>>> 
>>>> 
>>> 
>>> --
>>> Jeni Tennison
>>> http://www.jenitennison.com/
>>> 
>> 
>> 
>> ----
>> Ivan Herman, W3C
>> Digital Publishing Activity Lead
>> Home: http://www.w3.org/People/Ivan/
>> mobile: +31-641044153
>> GPG: 0x343F1A3D
>> WebID: http://www.ivan-herman.net/foaf#me
>> 
>> 
>> 
>> 
>> 
>> 
> 
> --  
> Jeni Tennison
> http://www.jenitennison.com/


----
Ivan Herman, W3C 
Digital Publishing Activity Lead
Home: http://www.w3.org/People/Ivan/
mobile: +31-641044153
GPG: 0x343F1A3D
WebID: http://www.ivan-herman.net/foaf#me