- From: <martin.hepp@ebusiness-unibw.org>
- Date: Fri, 2 May 2014 23:21:48 +0200
- To: Jason Douglas <jasondouglas@google.com>
- Cc: kevin.polley@mutualadvantage.co.uk, W3C Web Schemas Task Force <public-vocabs@w3.org>, Aaron Bradley <aaranged@gmail.com>, Jay Myers <jay.myers@bestbuy.com>, Mike Bergman <mike@mkbergman.com>
Hi Jason:
On 02 May 2014, at 19:03, Jason Douglas <jasondouglas@google.com> wrote:
> Could we use the existing extension mechanism instead of inventing a new one? Something like add productSpecification that expects QuantitativeValue and then extend it. So:
>
> {
> @type: Product
> productSpecification/screenSize : {
> value: 46
> unitCode: "CMT"
> }
> }
>
IMO, no. Of course, the meta-model for quantitative and quantitative values in schema.org, imported from GoodRelations, is powerful. But it is targeted at data of a cleaner state than the new proposal:
- You have to know whether the value is a quantitative, a qualitative, or a Boolean value. That is non-trivial and clumsy to do at the level of populating an HTML template. For instance, a string "12" may be a number or it may be a code. In fact when we wrote tools for converting vocabularies to OWL ontologies adhering to the GoodRelations meta-model, like http://wiki.goodrelations-vocabulary.org/Tools/PCS2OWL, the hardest part is to make this distinction.
- Also, http://schema.org/QuantitativeValue requires a degree of formality that is good for consumption, but hard for data publishers. You have to find the proper UN/CEFACT Common Code and split ranges properly.
- http://schema.org/PropertyValue is a less structured way of exposing the characteristics of an entity. I think that consumers of schema.org in RDF worlds will try to lift such data to http://schema.org/QuantitativeValue and http://schema.org/QualitativeValue and match the property to a standardized one. But that is another story.
We could have made http://schema.org/QuantitativeValue blurrier by adding unitText and propertyID to it, but after quite some thinking I have a strong opinion that we should keep the lighter structures in a dedicated type.
Martin
Received on Friday, 2 May 2014 21:22:16 UTC