A new music recommendation system from Sun
A music store like Apple iTunes contains more than 5 million songs today. And there are plenty of similar music stores online. With people posting online their own creations or excerpts of the concerts they attend, it's possible that a million new songs appear every day in a near future on the web. So how will you find new music you like? Right now, two approaches are prevalent: Amazon and other sites use collaborative filtering while Pandora and others use content matching. Both approaches are time-consuming, using both humans and computers. Now, according to Network World, Sun Microsystems is about to release an open source music recommendation technology far superior to current systems and totally automated. Read more...
This open source project, Search Inside The Music, has been led at Sun Labs by Paul Lamere. He followed several directions including 3-D visualization tools. As an example, you can see above "a 3-dimensional visualization that can be used to explore a large music collection." (Credits: Paul Lamere/Album artwork courtesy of Magnatune)
Now, why is the Sun's system better than existing ones? "Sun has built 'one of the best music similarity algorithms' that's based on the actual sound, with machine learning that analyzes features such as frequency and beats per minute to map out the rhythm structure, and determine the genre and which instruments are playing, Lamere said. Sun has taken advantage of prior research into speech recognition technology to tease out the features that correspond with the timbre of music and can be measured with computers, he said."
It's also much faster. In "Analyzing the tech biz Sun Micro spins its music software" (April 9, 2007), The Browser, a CNN Money blog, provides additional information , "Lamere says it takes a Pandora staffer about 20 minutes to categorize each song. There's a place for that, he says, but Sun's system is fully-automated -- it's a computer-driven algorithm capable of whipping along at a clip of '3 seconds per CPU per song.' The algorithm can’t tell the good from the bad, but it can sort the songs instantly by musical style."
But let's return to the Network World to discover that sound recognition technology is just one of this Sun's open source project. "Sun’s other innovation is a tagging system that categorizes music based not on who’s purchased it but on its attributes, described with tags like 'quirky,' 'indie,' 'rock,' 'fast,' 'frenzied,' '90s,' or 'cute' and 'fun.' Sun is compiling these tags by searching reviews, lyrics, music blogs, social tagging sites and artist biographies, and incorporating the information into a prototype search engine.
Here is an example of how the system works. "Querying Sun's prototype search engine for Led Zeppelin brings up a list of recommended artists -- or 'tagomendations' -- such as Pink Floyd, Queen, the Rolling Stones and Jimi Hendrix. The user can then click a 'why?' button to find out which tags overlap with the recommended artist and the one searched for. Hendrix is recommended for Led Zeppelin fans based on tags like 'guitar gods,' 'classic rock,' 'guitar virtuoso' and 'psychedelic.'"
For more information, be sure to read the whole Network World article. You also should read what David Berlind, a fellow ZDNet blogger, wrote about some cool search technologies in the Sun labs (December 13, 2006). Finally, don't miss Lamere's own blog, Duke Listens!. In his latest post, dated October 27, 2007, he wrote that he will start a new project. This is quite logical if the 'Search Inside the Music' is released as an open source project in the weeks to come.
Sources: Jon Brodkin, Network World, October 30, 2007; and various websites
You'll find related stories by following the links below.