Re: [TMO] coding XML patient data in the Indivo schema

> 2. I favor doing all significant transformations at the RDF level -- not
> XML.  In other words, transform to/from XML only at the edges, using a
> trivial transformation between XML and RDF, and then use ontologies and
> rules to transform to/from that RDF and RDF that is expressed in your
> preferred ontology, as illustrated on slide 21 of "RDF as a Lingua
> Franca: Architectural Strategies":
> http://dbooth.org/2009/stc/dbooth-SemTechConf-2009.ppt
>
>
absolutely - graphs extracted, then processed/enhanced, then serialized
selectively to the reports you need. BTW, the presentation is interesting,
using named-graphs to time-stamp "snapshots".