Ever since Twitter first opened its doors spam has been a problem on the network, and while their team has worked hard to battle this issue Twitter’s uniqueness as a system, plus its open API, has made this a particularly difficult problem to solve.
Enter BotMaker, a new system designed by Twitter that they say forms a solid foundation for defense against unsolicited content.
BotMaker is made up of three parts that break down the management of spam into smaller jobs. Scarecrow tries to stop spam tweets in real-time before they’re even published on Twitter by identifying dubious account names and URLs. Sniper takes on whatever Scarecrow missed by continuously classifying users and content. Finally, other models work periodically to track user behaviour over periods of time to make the overall BotMaker system smarter.
BotMaker already manages billions of events every day and Twitter says that they’ve seen a 40 percent reduction in spam metrics since it launched.
“BotMaker has ushered in a new era of fighting spam at Twitter,” writes Raghav Jeyaraman, anti-spam engineering lead at Twitter. “With BotMaker, Twitter engineers now have the ability to create new models and rules quickly that can prevent spam before it even enters the system. We designed BotMaker to handle the stringent latency requirements of Twitter’s real-time products, while still supporting more computationally intensive spam rules.”
(Spam image via Shutterstock.)