Warning - this applies to the Beta3 release of NGS server and later. RDFa will not work on the main site until sometime after the official NGS release --Kurtjx 21:11, 16 September 2010 (UTC)
RDFa is a W3C recommendation for embedding RDF data in HTML. The insertion of RDFa into MusicBrainz is part of the LinkedBrainz project and is intended to follow the NGS to RDF mappings.
There are several ways to consume RDFa present in MusicBrainz pages.
Using the rapper command line tool is perhaps the easiest way to get a feel for the model embedded in the MusicBrainz RDFa. On Ubuntu/Debian systems you can install with
sudo apt-get install raptor-utils
Then you can use rapper. Assuming you are familiar with the turtle RDF syntax, you can view the RDF model on a page using
rapper -i rdfa -o turtle MUSICBRAINZ_URL
Another option is to use the web interface at check.rdfa.info. However, MusicBrainz's restrictive robots.txt is obeyed by that interface so you must copy and paste the HTML source using the direct input option. The copy-and-paste workflow is quite annoying, but no software or libraries need to be installed.
A rather comprehensive list of libraries in various programming languages for consuming RDFa at the RDFa.info wiki.
Notes about pagination
Information about entities in the MusicBrainz website are generally spread across several tabs (tabination). Long lists of results (e.g. if an artist has several hundred releases) are spread across several pages (pagination). The RDFa in MusicBrainz adheres to this same tabination/pagination scheme.
Currently, the rdfs:seeAlso predicate is used to refer to other pages in the tabination scheme and the xhv:next, xhv:prev, xhv:first, xhv:last predicates are used to deal with pagination.