A couple of months ago, Tom Tague, who leads the Calais initiative at Thomson Reuters, presented at the New York Semantic Web Meetup. One of the projects he alluded to was announced today and reported in ReadWriteWeb: “Media Cloud Leverages Calais to Track News Trends“:
Media Cloud, a new project from the Berkman Center at Harvard University, has an ambitious goal: It will do the heavy lifting of analyzing stories from thousands of traditional news sources, analyzing the semantics of the content through Calais (covered here and here), and then providing tools to quickly get trending results.
The article also points to an interview of project developer Ethan Zuckerman by the Neiman Journalism Lab.
What particularly excites me about this project is the possiblity of comparing how different news organizations–or, better yet, different clusters of similarly biased news organizations–select and cover news. Ever since hearing Miles Efron present “The Liberal Media and Right-Wing Conspiracies: Using Cocitation Information to Estimate Political Orientation in Web Documents“ at CIKM 2004, I’ve been waiting for someone to take the next step and build analysis tools to compare the media “conspiracies”. For example, what stories are covered in the New York Times, but not in the National Review–and vice versa? Which details appear only in papers associated with one end of the political spectrum?
I don’t know that most people care about these questions. In fact, I suspect they don’t; my experience is that few people are interested in hearing viewpoints that challenge their own. But I fear that we are being personalized to death–that our control over what we read leads to the unfortunate behavior that we only let content through the filter if it reinforces our prejudices.
I know that Media Cloud won’t solve this problem on its own. But at least it’s a great tool for those who do want to broaden their perspectives, and I have hope that intellectually honest people will have the courage to learn from it.
