Orange Spain, a mobile operator, has two new music services – ‘New Talents’ and ‘Enjoy your Music’. Both are powered by Music Intelligence Solutions, a provider of AI (artificial intelligence) software for music recommendation and discovery.
Music Intelligence Solutions is an intriguing company. It was founded in 2002 and is based in Spain and the US West Coast. They describe themselves as “a diverse group of musicians and mathematicians”. Their customers range from record labels, to radio stations, to musicians and artists, to MP3 Player Manufacturers, to Music Retailers, to mobile operators. So they cover the whole spectrum of the music industry, which is unusual in a service provider.
Hit Potential
Their Hit Song Science technology aims to predict music chart success. It “overlays musical analysis with commercial data such as sales, highest chart position, and release date.” On their website aimed at music artists and producers, the technology is being promoted as a way to analyze the “hit potential” of music. You can read more about the technology here, plus The Guardian ran a good story about it back in January 2005.
In terms of how it is related to the Web, it is very similar to internet radio service Pandora – which analyzes music patterns, via software from The Music Genome Project.
The Orange “New Talents” service uses Hit Song Science. The idea is to allow independent artists to upload their new music onto the Orange website, providing them with a hit potential rating free of charge.
Recommendations
MI Universe technology, which powers Orange’s ‘Enjoy your Music’ service, is about giving music recommendations to users. It has two aspects to it – “soundalikes” and the “music taste test”. The aim is to match up music with “the user’s music personality”.
How does this work? It uses a clustering engine to find music patterns (including a “methodology for detecting social trends”) and a recommendation engine to do the matching with peoples tastes. This technology is being aimed at online music retailers, in-store terminals and mobile services. Here’s how it’s being used in the ‘Enjoy your music’ part of the new Orange Spain services:
“…when an Orange customer is browsing music on his phone or on the Orange website, he can click on a track he likes to generate “more like this”. The recommendation will be presented as a list on his phone, or graphically as a “mini-constellation” on the website. The customer will be presented with tracks from mainstream artists, plus tracks with a high success potential rating from independent artists, all of which he or she can sample and download. The tracks presented will have a high degree of sonic similarity with the initial track, and a strong correlation with the customers musical tastes.”
Summary
You can see the potential of music analysis software to power music services on the Web – for clustering, social patterns, user recommendations, personalization, and so on. Pandora is already proving it and these new services by Orange in Spain bring the experience to the mobile Web world. I’m expecting this to be a ripe area for growth in online music services – and also, in time, tv and video on the Web.
Originally published on ReadWriteWeb (archived copy)