Facebook on Friday said it has been trying to detect fraudulent activity and battle against spammers who promise to deliver "likes" to its members.
There are many companies that are promising to sell "likes" to Facebook members to make them appear more popular.
"We write rules and use machine learning to catch suspicious behavior that sticks out. When we catch fraudulent activity, we work to counter and prevent it, including blocking accounts and removing fake likes all at once," Facebook site integrity engineer Matt Jones said in a blog post.
"As our tools have become more sophisticated, we've contributed some of our spam-fighting technology to the academic community as well, in hopes of helping other companies combat similar problems."
Jones said that Facebook if necessary takes the spammers to court "to remind would-be offenders that we will fight back to prevent abuse on our platform. We also limit likes per account to make spammers' operations less efficient."
Facebook uses various techniques including algorithms to detect when there is a suspicious spike in "likes."