The general idea (and the reason for the name), is that RDF can provide the glue, or be the unknown missing thing (i.e. "x") that is needed in order to run experiments that combine hitherto incompatible silos of information in order to see if the unified view can reveal things that the isolated data cannot.

To do those experiments, it is necessary to build that help extract, transform and enhance raw data. We use the RDFX domain to publish experimental web services for such purposes.