infosthetics | ViralSearch [microsoft.com], developed by Jake Hofman and others of Microsoft Research, visualizes how content spreads over social media, and Twitter in particular.
ViralSearch is based on hundred thousands of stories that are spread through billions of mentions of these stories, over many generations. In particular, it reveals the typical, hidden structures behind the sharing of viral videos, photos and posts as an hierarchical generation tree or as an animated bubble graph. The interface contains an interactive timeline of events, as well as a search field to explore specific phrases, stories, or Twitter users to provide an overview of how the independent actions of many individuals make content go viral.
See also NYTLabs Cascade: How Information Propagates through Social Media for a visualization of a very similar concept.