The answer depends on how you "implemented RDFa". It is unlikely that the majority of your content is expressed as static information, so it is also unlikely that the majority of your content requires scraping.
There are tools, such as D2R Server, that give you facilities for exposing your underlying datastore as a read-only SPARQL endpoint. The only trick will be if you do have static content and wish to expose that as automatically generated RDF as well. That would require some finessing.