March 18th, 2009

Follow online media trends with Amber

Today, i’m happy to officially announce project Amber. In works over the last few weeks, it shows how to use Trident to add animations to an online interactive application that analyzes media trends. While previous demo videos showed some of Amber’s capabilities on test data sets, this high-definition video (watch in full-screen mode) shows Amber tracking the Digg tech news trends in real time.

To see Amber in action, click the WebStart button below and select one of the available Digg topics (last column):

Here are a few screenshots of Amber. The login screen (fades out on pressing the “Start” button):

The grid of story thumbnails (slide in and up):

Infotip showing story title and description (slides and fades in):

Another infotip, this time above the story thumbnail to fit in the frame:

Frequency chart of keywords in the shown entries – allows to analyze the popular topics:

The full code for Amber 1.0dev (code-named Ashberry) is available in the SVN repository, and Rémy Rakic (@lqd on Twitter) has already created a Twitter provider that allows analyzing public timelines:

To create a custom provider, take a look at the org.pushingpixels.amber.digg.DiggProvider class. One of the next blog entries will show another provider that uses the NY Times online APIs.