Schema.org models

Hello, everyone. I'm a product strategist with Intel's Cloud Service Platform group. We're building a platform of cloud services that enables developers to write cross-platform, cross-garden apps. Our flagship product is Intel Identity Services, but the suite of services includes Location, Context, and Commerce, Recommendations, Catalog, and Curation.

I am experiencing the need for an open, platform-agnostic way of representing common objects such as books, movies, restaurants, and so on. It turns out that Schema.org addresses many of these needs... but with a firm linkage between the object models and their HTML representations.

However, when you look at the underlying ontology/models that Schema.org implements, there is no reason that the object models and HTML representations need to be so welded together.

I am looking at adopting the object models for data interchange between applications and even between developers, relying on the underling OWL object descriptions but representing the data in HTML, XML, JSON, or other formats as appropriate (including storing in databases with schemas mapped to the Schema.org OWL definitions).

Benefits to developers include ease of interchange between applications, ease of repurposing data for Schema.org-marked-up web pages, and a large number of well thought out object models.

Does anyone have experience with this kind of application of the Schema.org models, or feedback/advice on this approach?

-Brooks