Twitter cuts spam by 40 percent
Twitter has been able to reduce the number of spam messages on Twitter by 40 percent with a new tool. Some of the spam is already noticed and blocked during posting, but existing spam attacks are also reduced faster.
With the tool Twitter employees can easily add new anti-spam measures. It used to take hours or even days as a code change had to be made, tested, and then rolled out. With Botmaker, as Twitter’s new tool is called, this can be done within minutes. In fact, Botmaker consists of a large number of small bots that search for spam messages.
Botmaker has its own, easy-to-read syntax, in which conditions and actions must be drawn up. Each bot consists of a condition and an action. For example, a bot can search for a tweet with a certain text or URL and then delete it.
Botmaker already works while posting a tweet. This should lead to spam being blocked even before users can come into contact with it. However, more complex bots only run afterwards, to avoid too much latency when posting a tweet. There are also bots that look at the behavior of a user over a longer period of time.
Botmaker is now Twitter’s most important anti-spam tool, the company states itself, and the tool is also used on other systems. It is unclear for what further purpose the tool can be used. Nor is it clear whether Twitter has plans to make Botmaker open source.