Meme-tracking and the dynamics of the news cycle
Jure Leskovec, assistant professor of Computer Science at Stanford University
Tuesday, February 16, 2010 at 12:30 pm
Berkman Center, 23 Everett Street, second floor
The news cycle --- the set of temporal patterns by which news grows and fades over time --- has increasingly come to be seen as an integral part of public discourse and the political process. But despite extensive qualitative research, there has been very little work studying the properties of the news cycle at a quantitative level.
I will discuss our analysis of approximately 1.6 million mainstream media sites and blogs for a period of three months, covering about 1 million articles per day. We developed methods for tracking quoted phrases through this collection of articles, using a combination of techniques based on graph partitioning and sequence alignment. We found that tracking such phrases provides a level of resolution capable of exposing novel and persistent temporal patterns in the news cycle. In particular, we observed a typical lag of 2.5 hours between the peaks of attention to a phrase in the news media and in blogs respectively, with divergent behavior around the overall peak and a ``heartbeat''-like pattern in the handoff between news and blogs. We also developed and analyzed a mathematical model for the news cycle that captures some of the dynamics we observe.
This is joint work with Lars Backstrom and Jon Kleinberg.
Jure is an assistant professor of Computer Science at Stanford University. His research focuses on mining and modeling large social and information networks, their evolution, and diffusion of information and influence over them. Problems he investigates are motivated by large scale data, the Web and on-line media. He received of three best paper awards and a ACM KDD dissertation award, won the ACM KDD Cup in 2003 and topped the Battle of the Sensor Networks 2007 competition.