Derivadow.com

Over the last few weeks we’ve been working with Rob Kaye to integrate the MusicBrainz database into the BBC’s music site. And we’ve just integrated MusicBrainz to provide discographies and track listings across our 700 or so artist pages. Well I say we, but I really mean Michael and Stephen who have done most of the work - so a big thanks to them.

This is an exciting project. MusicBrainz is an open source, community maintained database of music information. We’re using MusicBrainz to provide centralized Music IDs for artists, albums and tracks to underpin our move to a data-driven music offering. And we’re starting with discographies (which I guess in itself isn’t exactly exciting - but little steps and all…)

Providing a centralized database of Music IDs is important because it means we can uniquely identify all our music information. When I use iTunes their database has no idea how, or even if, different versions of a piece of music are related. For example, it doesn’t know if a song on two different albums are different versions of the same song or if its the same song released on two albums. This is a bit rubbish because it means I can buy the same song twice. Apple don’t want you to do this and indeed they do stop you buying the same track on the same album twice; but not the same track from different albums. If Apple could uniquely identify tracks that wouldn’t happen, but because they don’t have centralized Music IDs they can’t.

MusicBrainz provides unique IDs, plus a bunch of useful information (metadata) about music. This is pretty cool because eventually it could mean we can use it to unambiguously identify all the music played on the BBC. Although for now it means we can publish discographies and track listings.

These IDs are being licensed to the BBC, by MusicBrainz, under a creative commons license which means they can be shared, reused and remixed. So a bit like ISBNs or post codes (zip codes), MusicBrainz IDs provide a universal and world-unique identifier for music so we and others can identify, link to, reuse and map the world of music both across BBC sites and between different organizations who also use MusicBrainz.

And because we’ve also released our associated album reviews under a creative commons license. And formatted the data using the hReview and hCard microformats (microformats are a set of open data formats built upon existing, adopted standards) others will find it easy to remix our content.

So what might all this mean for the future? Well last weekend’s London Hackday saw a tantalizing glimpse of one possibility with Michael’s clickable track listings prototype. The data is taken directly from our broadcast systems and matched to MusicBrainz to dynamically generate:

• station aggregation views aggregating the artists played on each station into both charts (most played that day, week and since the prototype started running) and artist clouds;
• programme views which are similar to the station aggregation views but for each programme and links through to each episode;
• episode views including track listings with links through to artist pages;
• artist pages with biographies pulled in from Wikipedia (MusicBrainz include links to Wikipedia) and links back to the programmes that have featured that artist.

This is all possible because MusicBrainz is being used to tie it all together - because it provides canonical, addressable Music IDs that allow each artist to be uniquely identified - no matter which station or programme they were played on.