Based on its original concept, FairSpin for Twitter offers followers a chance to filter their news tweets and follow three categories: left-leaning, right-leaning and neutral news. The site's original dashboard also has a Twitter page specific to left, right and neutral tweets.
While the concept of tracking media bias is certainly not new, the methodology to determine it has varied across services. FairSpin pre-assigns media bias ratings across major outlets and only aggregates news from a select few sources that appear on political aggregation site Memeorandum. Those ratings are then further voted on by members after they've been published.
Meanwhile, competitor Skewz, allows members to upload any news and reveal their own political agendas while voting on every article. Similar to Firefox plugin SpinSpotter, news sources can include everything from small blogs and videos to major traditional outlets.
And finally in 2008, Microsoft's Natural Language Processing group spoke openly about their research project BLEWS. In this project the group categorized a random sampling of blog posts and weighed words for their emotional charge. For example, words like "failure", "progress" and "strong" indicated considerable bias. From there a computer was taught the words and automatically determined the blog's political partiality and emotional charge.
It's important to note that while a journalist's agenda and machine linguistic learning are well-known factors in determining bias, other well-established points of reference include an outlet's connection to advertisers, government, PR agencies and corporate-sponsored media speaking engagements. For more on media bias methodology, check out the Global Media Journal. (Note: In the spirit of deconstructing the media, we're not saying this is a unbiased document either, it's just a good overview of the topic)