Track 3: desktop, Desktop – Google 2007 JavaOne Advance Conference Guide User Manual

Page 25

technical sessions | track three : desktop

desktop

TS-1548 Search Inside the Music: using Signal Processing,

Machine Learning, and 3-d visualizations to discover
New Music

Paul Lamere, Sun Microsystems, Inc.

This presentation discusses Search Inside the Music, a Sun Laboratories
research project that is exploring new ways to help people discover new
music even as our music collections get very large.

As online music collections grow to many millions of songs, finding a new
song we might like is becoming very difficult. The Search Inside the Music
system can help you find new music by finding music that sounds like
music we already know and like.

SITM, written entirely in the Java programming language, uses digital
signal processing and machine learning algorithms to build a music-
similarity model that can predict how similar or dissimilar a pair of songs
sound. SITM uses this model to recommend music by finding music that
sounds similar to music you already know and like.

Not only can you use this music-similarity model to help recommend
music but you can also use the model to generate a more engaging,
immersive interface to your music. SITM uses the music-similarity model
to generate a “music space,” a 3-D representation of a music collection,
in which songs are positioned according to music similarity. In this music
space, classical music may be clustered in one corner, trying to stay as
far away from punk music as possible, whereas blues finds a home near,
but separate from, jazz and rock. This visualization encourages music
exploration. You can audition new songs by clicking on a song in the
visualization and find similar-sounding songs by clicking on a song’s
neighbors. You can generate interesting playlists by creating paths
through this music space.

The presentation discusses some of the problems inherent in traditional
music recommenders and how a content-based approach to music
recommendation can help improve music recommendations. It covers
some of the algorithms involved in building a music-similarity model,
including the digital-signal-processing algorithms used for extracting
music features and the machine-learning algorithms for identifying
significant patterns in music. It also discusses some of the algorithms
used to generate immersive interactive visualizations of a music space,
using the Java 3D API. It concludes with a demonstration of the Search
Inside the Music system.

dESKToP

Java technology on the desktop is more vibrant
than ever. Rich, interactive desktop applications
and applets based on Java technology play an
increasingly important role in delivering and
supporting standard applications as well as
new innovations. Desktop development in Java
technology continues to thrive on all major
desktop platforms, including Macintosh, Linux,
and Windows. Come learn about technologies and
techniques in the following areas:

• Swing: Making GUI application development

easier and more powerful

• Making GUI applications more usable,

appealing, and dynamic

• GUI and graphics-related technologies
• 2-D and 3-D graphics and imaging
• Deployment: Making your standalone, browser,

and Java Web Start software applications
available to the masses

• Cool stuff that shows innovation on the desktop

TS-1548 Search Inside the Music: Using Signal Processing, Machine Learning, and 3-D

Visualizations to Discover New Music

track three : desktop

TS-2901 A Real-Time Garbage Collector for a Real-Time Java Virtual Machine

TS-2906 Garbage-Collection-Friendly Programming