Normalization: A Gracenote

Normalization of asset discovery data is not easy to explain, but I ran a across a really great analogy when downloading music from some CDs to my iPod. As every iTunes user knows, when one sticks a CD into their PC to transfer of music to their iTunes library, a service called Gracenote kicks in, identifies the CD being transferred and conjures up the album name, track names, artists, and music genre, and it includes with the music files stored on the iPod.

In doing this, Gracenote performs both asset discovery and data normalization, very much akin to what we do at BDNA for enterprise technology. Asset discovery kicks in when Gracenote recognizes the CD and zooms off to retrieve artist, track and timing info from its database. Gracenote also performs a fair amount of data normalization, particularly with multi-artist, compilation CD sets.

But here, it might be more accurate to say that Gracenote struggles with normalization as it can return some very inconsistent information, with multi-CD content proving particularly troublesome. There are some multi-CD sets that Gracenote handles with no problem. Richard Thompson’s “Watching the Dark” three-CD retrospective appears on my iPod exactly as it does on its native CDs. Others are not so good. Eric Clapton’s “Crossroads” shows up as seven albums, not four.  Gracenote was thrown by the first CD, classifying Clapton’s work with The Yardbirds, John Mayall, John Mayall and the Bluesbreakers, and Cream as four separate albums.  It also has problems with records that have the same name. Both Rod Stewart and Traffic issued greatest hits collections called “Gold,” and tracks from both CDs are mixed up together under the “Gold” title on my iPod.

Don’t get me wrong. I really like Gracenote and wish them well. I am acting as a friend when I say that it would be even greater if they could do becoming more reliable in accurately and consistently in classifying and reporting track information to my iPod. In other words, doing a better job of normalizing the data it discovers.

I realize that my guilty pleasure-dominated musical tastes  betray me as a man of my generation. But sometimes I think that if this IT thing doesn’t work out, I might have future in the music business. But every time I mention this, I get the same advice: “Don’t quit your day job.”