search

schemaorg / suggestions-questions-brainstorming

Suggestions, questions, and brainstorming
20 stars 15 forks source link

Why not RDF? #272

Open krabina opened 1 year ago

krabina commented 1 year ago

I haven't found any answer to this yet, neither here nor in the mailing list.

While I perfectly understand the goals and importance of the Schema.org vocabulary, I want to know why Schema.org requires Microdata, RDFa or JSON-LD formats and does not acknowledge RDF (RDF/XML)?

The reason why I am asking this is that I work a lot with Semantic MediaWiki which has the capability to provide a RDF/XML-representation and to re-use schema.org vocabulary out of the box.

However, the Schema.org validator will not find any elements: https://www.geschichtewiki.wien.gv.at/Hedy_Lamarr even thought the source code of the page leads to the RDF representation: https://www.geschichtewiki.wien.gv.at/Spezial:RDF_exportieren/Hedy_Lamarr with the html code

Were there any discussions about the formats? I would be interested why the (longer existing and quite sophisticated and well-known W3C standard was not taken into consideration.

Is there a chance it will in the future?

stain commented 1 year ago

I don't think your question relates to the schema.org vocabulary but how the schema.org Markup validator can't find your RDF through content negotiation.

As your page do not contain any such markup, then it correctly shows that no schema.org is found in the HTML on that page.

You do have this tag which perhaps you expected the markup validator to follow?

<link rel="alternate" type="application/rdf+xml" title="Hedy Lamarr" href="/index.php?title=Spezial:RDF_exportieren/Hedy_Lamarr&amp;xmlmime=rdf"/>

It is unfortunately not common for RDF processors to follow such alternate signposting. Try for instance on http://rdf.greggkellogg.net/distiller which would try to content negotiation for application/rdf+xml but your server do not support this either.

The best option if you want visibility by search engines, who do not want to heuristically navigate to parse the many different RDF formats, you may embed the JSON-LD in the HTML using a <script> tag as documented by the JSON-LD standard.

If I feed your RDF/XML into the RDF Distiller I get quite a verbose JSON-LD (due to the many vocabularies and multiple redefined OWL properties).

Feeding this into the validator will show you that the validator is able to understand the additional properties, but of course it has no idea about their validity and the user interface is not designed for handling generic graphs that don't use schema.org.

You can then, if you navigate deep enough, find potential warnings etc such as:

image

krabina commented 1 year ago

My question was more about the general approach. I don't understand why schema.org only allows three technical formats for the vocabulary to be delivered, when there would be RDF/XML as at least equally powerful format.

But thank you for your detailed analysis, this is very helpful. Actually there is content negotiation, so we should probably point search engines to this link instead: https://www.geschichtewiki.wien.gv.at/Special:URIResolver/?curid=30218