Ontology-based semantic rendering is performed by retrieving a rendering script from a location represented by an ontology reference of a concept of content included in a document, the document further including an ontology dataset identifier of the concept and the ontology reference, and rendering the concept of content according to the rendering script and a mode of rendering represented by a rendering device.