Are President Barack Obama and Governor Mitt Romney sticking to their guns, or are they being inconsistent about certain issues in the run-up to the U.S. presidential election?
A semantic search company, Atigeo, has developed red2012blue.com and blue2012red.com, new websites for voters to learn where their favored candidates stand on the issues based on an analysis of their social media streams.
Using the visualization tool, you can click on an issue (“national defense,” “economy,” and so on), which will pull up potential inconsistencies. It’s a machine learning tool, so presumably it will get smarter the more times that people use it.
View a video demonstration here.
The data is drawn directly from the following feeds, since before January 1:
- Democratic National Committee: Twitter, RSS feeds, and Facebook page
- Republican National Committee: Twitter and Facebook page
- President Barack Obama: Twitter feed, RSS feed and Facebook page
- Governor Mitt Romney: Twitter feed and Facebook page
Atigeo is one of the many analytics companies that can sort through large stores of unstructured data, like tweets, posts, and email. A few have begun monitoring the exploding volume of social media content produced by politicians’ communications teams in an effort to understand how they are engaging with voters.
According to Atigeo, of Facebook’s 995 million users, more than 27 million have “liked” President Obama, while the Democratic Party has approximately 400, 000 users who are “liking” their fan page. The GOP has greater than 500,000 users “liking” their fan page, and Romney’s page has more than 3.8 million users who have “liked” him.
Note: It’s unlikely that these fans and followers are all genuine (read up on recent research findings that a portion of Romney’s Twitter following is likely fake).
The company claims to be a “compassionate” Big Data vendor. Analytics providers are probably about as compassionate as your average politician, but to be fair, the service is useful and available for voters to access for free. It’s a fun tool to play with — let us know what you find out!
VB's research team is studying web-personalization... Chime in here, and we’ll share the results.