If you have used Twitter, you either love it or hate it.   Regardless of how you feel on the subject, social media sites, like Twitter, have become a source of news for some, though if you think people only watch news on television they already agree with, you can imagine what their Twitter feeds are like.

John Timmer at Ars Technica discusses some Yahoo! researchers who say they can create an automated system that identifies newsworthy events and judges their reliability with an accuracy of nearly 90 percent.

They used Twitter Monitor to see what was trending and gathered all the tweets on thousands of topics - then they used humans to help calibrate 'credibility' (caveat emptor on that) and used support vector machines, Bayesian networks, and decision trees to help them determine how to find the most credible sources.

Which method worked best? Read  Accurate and credible news tweets? Automated system finds them by John Timmer