[In response to recent posts about using RDF, I have some comments from Hoylen Sue at the DSTC at UQ, Brisbane] Thomas Beale writes: > people keep suggesting RDF as a better way of modelling classes in XML. > What do you think? Where does XML-schema fit in this debate? RDF is a framework for metadata, consisting of both a model as well as a syntatical representation in XML. It is a key part of the Semantic Web initiative. It is being often sold as something that is at a higher level than XML. I think people keep suggesting it because they believe there is value in complying to standards where possible, but I don't think the RDF world is mature enough at this stage to provide much tangible benefits from using it. Although a clinical record could be considered as clinical data (the metadata) about a person (the resource) and fitted into RDF, you don't gain much benefits from doing so. It might be something we could do for the XML instances, just to claim RDF compatiability. If trying to fit the EHRs into the RDF model is trivial then let's do it, but I suspect some of the more complex structures will not work. In the XML Schema debate, RDF is a totally different beast. Currently RDF is not about modelling the structure. There is a old RDF Schema proposal that has been dormant for quite a while, but it is very limited compared to XML Schema and there are talk (by some, but not all) about moving it towards using XML Schema rather than coming up with a separate schema just for RDF. A RDF core working group has just been started in the W3C to fix up and build on the existing RDF work. What are the arguments for using RDF and RDF Schema? Hoylen P.S. XML Schema has just become a full Recommendation (on Wednesday 2nd of May 2001).
