Re: Schema.org Extension Proposal for Automotive (and other vehicles)

Hi Vicki:
On 01 Aug 2014, at 17:20, Vicki Tardif Holland <vtardif@google.com> wrote:

> I have a quick question as I skim through.
> 
> It looks like the MotorizedRoadVehicle allows people to model an automobile model in general (e.g. Honda Civic) as well as a specific vehicles (e.g. my Honda Civic). Is that correct?
> If so, does that model apply to the other vehicle types?

Yes, you can model both individual cars and car makes and models.

This comes from the GoodRelations model for products. In principle, you have three subclasses of Product in schema.org (and GoodRelations):

a) IndividualProduct: A real product (my Macbook)
b) ProductModel: An abstract ideal of a product, similar to a datasheet or model
c) SomeProducts: A bag of multiple, real products

Let's put c) aside for the moment.

a) would be a real car (my Ford T)
b) would be the blueprint of a car (the Ford T)

a) and b) are necessary and useful, because we can define properties for b) and, once me know that a product is of a certain make and model (either by a direct link or on the basis of strong identifiers like GTIN, UPC, ...), we can pass along all properties from the model to the individual product. Only real individuals e.g. have serial numbers.

If a site owner is not able to make that distinction, he or she can simply use http://schema.org/Product.

c) is meant for modeling new products in shops, where the publicly exposed entity is typically not a real individual, but only an existentially quantified individual. E.g. the URI of a book in a Amazon book page about a certain book will not identify a single copy that will be gone when sold.

Now, once you add classes for types of products, like "Car", "Book" or "Bike", the idea in GoodRelations is that you simply use multi-typed entities. E.g., in Turtle notation_:

myCar a schema:Car, schema:IndividualProduct.

CarModel a schema:Car, schema:ProductModel.

The rationale is that otherwise, you would have to create three classes for each specialization of the product class (e.g. CarModel, IndividualCar, SomeCars; VanModel, IndividualVan, SomeVans; etc.).

If the consuming client (e.g. Google) is able to infer from the context whether an entity is a model or individual, it is fine if sites just use schema:Product or schema:Car or schema:Van. Personally, I am convinced that having the distinction is valuable, but I admit that it is one of parts of GoodRelations and schema.org that are not widely understood and properly used.

We also have to admit that historically, GoodRelations was meant as a conceptual model for publishing AND consuming data. In schema.org the conceptual structures used for publishing can be weaker, because the consumers could lift the data to more detailed internal data structures. 

I hope that helps.

Martin

> 
> Thanks,
> Vicki
> 
> 
> Vicki Tardif Holland | Ontologist | vtardif@google.com 
>  
> 
> 
> On Thu, Jul 31, 2014 at 4:58 PM, mfhepp@gmail.com <mfhepp@gmail.com> wrote:
> Dear all:
> 
> We have just finalized a proposal for adding better support for cars (sales and rental) and other vehicles to schema.org.
> 
> Please find the proposal in the following places:
> 
> * Git repository
>     https://mfhepp@bitbucket.org/mfhepp/sdo-vehicles.git
> * W3C Wiki
>     https://www.w3.org/wiki/WebSchemas/Vehicles
> * Development version of schema.org including the proposal
>     http://sdo-vehicles.appspot.com/
> 
> See e.g.
>     http://sdo-vehicles.appspot.com/Car
>     http://sdo-vehicles.appspot.com/Van
>     http://sdo-vehicles.appspot.com/Motorcycle
> 
> For a quick overview of the proposed changes, see
>     https://www.w3.org/wiki/WebSchemas/Vehicles#New_Elements
> 
> Essentially, the proposal adds the most relevant parts from the following vehicle-related extensions for GoodRelations:
> 
> - Vehicle Sales Ontology (VSO), http://purl.org/vso/ns
> - Volkswagen Vehicles Ontology (VVO), http://purl.org/vvo/ns
> - Used Cars Ontology (UCO), http://purl.org/uco/ns
> 
> Any feedback is very welcome.
> 
> Best wishes
> 
> Martin and Karol Szczepański
> 
> -----------------------------------
> martin hepp  http://www.heppnetz.de
> mhepp@computer.org          @mfhepp
> 
> 
>

Received on Friday, 1 August 2014 21:25:57 UTC