Data includes posts on Twitter between Oct. 1 and Dec. 2 from more than 1,700 accounts identified as fake or exhibiting coordinated or bot-like behavior.