TWASE is an anti-spam engine that given a Twitter username can determine whether the thing behind it is a spammer, marketeer, bot or just disruptive member of the Twitterverse.

We call these irritating accounts 'twam', or TWitter spAM.

Our technology is served up via an API so that developers might include our service in their own projects and applications.

To view a quick outline of TWASE, check out the 15 min platform overview

TWASE is still a beta effort and is nearing 1.0 after almost a year of spare time and on + off development, so please do mind your step as you use the platform and provide some feedback to help improve the Engine - see Contact.

To see an example of the return XML using a public demo Key, try this:

http://api.twase.com/d3m04p1k3yf0r73571ngpurp05350nly/latest/scan/all/jchivers

To start using the API, you'll need your own Key - see Keys.

Once you have a Key, read through the Usage section to see how to start using the platform.

The TWASE architecture processes many attributes pertaining to Twitter users; amassing data and generating peripheral toolsets such as Bayesian filters, statistical lookups, outlier detection and algorithms for delta value detection - in short, the Engine builds models of both legit Twitter users and twam accounts then comparing these models against a supplied user to test.

Eventually, TWASE will also enable developers to set customised filtering preferences such as follower⇔following ratios, tweet content regular expressions, and other cool things.

• desktop Twitter client - Help protect your users by canning unwanted @reply's and automatically blocking spammers on your user's behalf

• data mining app - Filter out undesirable content from your Twitter feed/stream or search results when processing tweet or Twitter user data

• user ranking and statistics - Discard bogus users from your analysis and focus on only genuine Twitter users

• existing platform - vet users when opening your existing platform to access from Twitter user accounts