Curating the Twitter Election Integrity Datasets for Better Online Troll Characterization

In modern days, social media platforms provide accessible channels for interaction and immediate reflection of the most important events happening around the world. In this paper, we, firstly, present a curated set of datasets whose origin stem from the Twitter's Information Operations efforts. Secondly, we analyze how troll activity fluctuates over time, and how it compares to a control group of real and active users. We present baselines for such tasks and highlight the differences there may exist within the literature (e.g. [2]). Finally, we utilize the representations learned for behaviour prediction to classify trolls from "real" users, using a sample of non-suspended active accounts.