The millions of tweets have already been analyzed for various events and characteristics, ranging from football, basketball and music television awards, to the mood of people, their morning greetings, successful flight arrivals and the weather.
Now The New York Times has given in to the pressure, and released a view into the buzz surrounding the mid-term elections in the US on Tuesday. Their interactive bubble chart [nytimes.com] highlights the number of posts related to candidates running for governor or the Senate. The size of each bubble relates the number of Twitter messages from and directed at each candidate's account, and the number of tweets for each candidate that got retweeted by other people.
I am not sure what the spatial location of each bubble actually denotes, although the chart, on a purely visual level, does remind me of those generated by the multi-dimensional scaling technique, which should normally put candidates that 'are' similar close to each other.