LinkedBrainz Summary

user warning: Unknown column 'captcha_type' in 'field list' query: SELECT module, captcha_type FROM captcha_points WHERE form_id = 'user_login_block' in /var/www/drupal/sites/all/modules/captcha/ on line 64.

Title of Primary Project Output: LinkedBrainz Live!

The type of user this product is for: LinkedBrainz Live is for application developers and institutions who require or use information about recorded music. Especially if thes require to interlink datasets or benefit from inference via automated reasoning.

Description of prototype/product: Our product is three-fold. Via MusicBrainz we now provide RDFa content on each of their pages which enables third party developers to access music metadata content in a machine readable format from the browser. We provide a dump the complete public MusicBrainz metadatabase in RDF such that any institution can get the dump and integrate it in their Linked Data store. Finally the project provides a SPARQL endpoint, i.e., a service to query the MusicBrainz linked data.

Link to working prototype/product:

Table of Contents to Project Blog (with diagrams):

The natural starting point for the project - for the purposes of making Linked Data content out of the MusicBrainz dataset, which is made available as a relational database under the Next Generation Schema (NGS) - is to use D2R.

D2R provides dynamic RDF, HTML and a SPARQL endpoint from the relational database, and allows a full dump to RDF to be run, according to mapping rules. Since MusicBrainz already provides HTML content for their metadata, this facility was not required.

In order to deal with the UUIDs used in the postgresql MusicBrainz database, a fork was made of the D2R Server. Downloading and installing this, and setting it up against the MusicBrainz dump is described in

Translation rules in D2R are based on the D2RQ RDF schema. LinkedBrainz operated a Wiki to discuss these rules, as documented in

One of the most important tasks for the LinkedBrainz project was to agree a canonical URI scheme, respecting the (in)famous httpRange-14 issue. This scheme was documented in

Recent updates to the Linked Data Principles (note the change from "RDF" to "RDF*" in Principle 3) have responded to the use of the RDFa standard for the exposure of Linked Data. Apart from its role in annotating the human-readable text, this has increasingly found acceptance in upgrading existing infrastructure to account for Linked Data content (especially where the redirect-based solution to httpRange-14 is unavailable). LinkedBrainz has focussed on the use of RDFa, then, in the exposure of live data from MusicBrainz, and on 19th May RDFa was rolled-out in every (content) page, to every requester, as documented in

The size of this huge contribution to the Linked Data Cloud was documented in

The first major use of the Linked Dataset was made at the ESWC Summer School, where students developed prototypes using the live RDFa and also experimented with the SPARQL endpoint provided by D2R, as documented in

Since, however, the scalability of the querying over dynamic translation was poor, the project has focussed on the provision of an effective dump, via the D2R mappings. It's announcement, and instructions for loading into triple store for better query performance are documented in

Step-by-step instructional documentation for reusing prototype:

Link to code repository and published data:, and

Link to technical instructional documentation: LinkedBrainz RDFa code follows but this does not lead to a dumpable documentation.

Licence used for code:

License used for data:

Licence used for (blog) content:

Project start date: 1-Aug-2010

Project end date: 31-Jul-2011

Total grant: 94894GBP

Project team:


JISC Funding Stream: