Twitter spammers under attack

 

Gallery

By Stewart Mitchell

Posted on 14 Aug 2009 at 14:58

Twitter users fed up of mass marketing and phishing attacks from supposed "followers" have a new tool in their defensive arsenal with Twitblock, which the developer claims can drastically hamper computer-aided accounts.

According to the company, TwitBlock is a junk filter for the micro-blogging service and uses a points-based ranking system to identify potential spammers and scammers.

The service combines user feedback and heuristics to allocate a score for potential miscreants. Points are earned for discrepancies such as missing profile information, computer generated user names and identical profile pictures, as well as how many subscribers have blocked followers.

Twitter exploits have continued apace in the wake of the recent DDoS attack that left the service offline for two hours, and rumours persist that the service is used to control botnets.

Sometimes real people and legitimate companies will display the same characteristics (as spammers.)

TwitBlock currently has a minimalist 234 users, yet knows of more than 10,000 blocked users and claims accuracy will grow with more subscribers.

However, regular Twitterers have complained that they have been caught up in TwitBlock's cull. "Sometimes real people and legitimate companies will display the same characteristics [as spammers]," the company says. "But we intend for the scanner to become more accurate. We also intend to offer 'whitelisting' so companies can apply to become exempt from showing up in scans."

Subscribers can currently flag followers as "not spam" to make sure their messages come through unmolested.

Twitter also announced that it will fold the phenomenon of retreats into its official feature list.