Files
Abstract
In the age of information, social media plays a vital role facilitating information flow. Unfortunately players with political agenda intervene and affect the information flow by spreading disinformation, that in turn affects our day to day life considerably. One recent such incident is the use of trolls - players who are paid to spread disinformation by agencies with intention to meddle with political events - that has affected the information flow significantly. This thesis focuses on a crucial problem of identifying the trolls, especially on twitter, from regular accounts by analyzing the user behavior. To the end, we also propose a pipeline to identify tweets that may potentially belong to a troll account by classifying how likely the tweet is a misinformation.