The sheer volume of media coverage of the American election cycle makes it impossible to keep on top of everything being reported, and even harder put it all together into a database worth analyzing from a sociological standpoint.

Thankfully, computers are excellent and chewing through immense mounds of repetitive information, and Election Watch from the University of Bristol does just that. It automatically trawls through thousands of articles, identifies actors, actions, and if the link between is a positive or a negative.

For the 2012 election cycle, more than 90,000 articles have already been parsed, and that's just the beginning. This tool is set to be presented at the 13th conference of the European Chapter of the Association for Computational Linguistics, and not only will they be showing off the data from this year, but also the last five election cycles (though that information comes strictly from the New York Times).

So, why is any of this useful — apart from making pretty graphs? On an academic level, it gives policy wonks a huge corpus of data to play, to rapidly track statements and shifting statements, without needing to hand code hundreds of articles. It also serves to create maps exploring the labyrinthine interactions between players on the political scene, where their views interact, and where they differ.