TweetBlocker by HashRocket is a great new web app that attempts to clean up your Twitter stream of all that spam. TweetBlocker uses a spam filter algorithm to grade each of your followers using the ABCDF grading system. You then have the ability to browse your followers based on grade score and select followers individually or in bulk and with the click of button – block them! However the unfollow feature that appeared when it was first launched a month ago has since been removed.

How it works

TweetBlocker uses Bayesian spam filtering as well as other criteria to grade each account including account age, how active the account is, and friends to followers ratio. However I was able to identify quite a few false positives in the A-D categories but the F category seemed to be dead on with only about 50 users. Now, I am sure I am followed by more than 50 spam accounts but I was satisfied with the F grade that these accounts received and proceeded to block them with ease.

The first time you sign in with your Twitter account you could expect to wait a while for the system to load and analyze all of your followers. It initially took about an hour to load 17,500 followers but was much faster each subsequent time after my first login.

Once your followers are loaded you can browse based on name, score, or grade. Hovering over an avatar will display the users profile along with their last three tweets.

On their homepage they have a  bookmarklet that makes it very easy to report a spammer and you can check the grade of any twitter user without logging in. They also have an API for third party developers to integrate TweetBlocker into their own applications. Very cool!

Conclusion

I suggest adding TweetBlocker to your arsenal of Twitter tools. Just be careful of who you block so that you don’t accidentally block legitimate friends or followers!