Algorithm identifies antisocial web users with 80% accuracy

Researchers studied 40 million posts made by 1.7 million web users. From this they could identify so-called Future-Banned Users (FBUs) - and Never-Banned Users (NBUs).