Ignore non-schema properties

Hi,

First let me introduce myself. My name is Dimitri van Hees and I am currently working on a research for the Dutch Government about publishing geospatial information on the web. It's called the Geo4Web Testbed and is closely related to the W3C groups 'Spatial Data on The Web' and 'Data on The Web'. I am involved in Research Topic #3: Crawlable geospatial data using the ecosystem of the Web and Linked Data (http://github.com/geo4web-testbed/topic3). For this purpose we want to make use of the Schema.org geospatial classes and properties to investigate if Search Engines will crawl the geospatial information.

Here is my question. Apart from multiple other representations, I'd also like to expose a JSON-LD representation of my resources. JSON-LD is fine with multiple vocabularies and you can in fact 'ignore' properties by not defining them in the @context. This means this is my source (note that I shortened the polygon value for readability):

{
  "@context": {
    "geo": "http://schema.org/geo",
    "polygon": "http://schema.org/polygon"
  },
  "@type": "http://schema.org/Place",
  "ID": 1,
  "MUNICIPALITY": "Leeuwarden",
  "WATER": "NO",
  "geo": {
     "@type": "http://schema.org/GeoShape",
     "polygon": "53.24330130 ... 2536921"
   }
}

>From a Linked Data and JSON-LD point of view this is completely correct. The properties ID, MUNICIPALITY and WATER aren't mapped in the @context so the parsers ignore this. However, the Structured Data Testing Tool reports errors that those properties aren't known as properties of http://schema.org/Place. The same errors occur when I map those values to another (my own) vocabulary.

I was wondering if this really is an issue for Google or is this just an issue with the testing tool?

Looking forward to hear from you!

Kind regards,

Dimitri

Received on Friday, 15 January 2016 12:49:47 UTC