Sentiment Analysis to Identify Patterns in Online Aggression

Abstract

Rise of internet has given a lot of worth and power to the voice of common man and it won’t be wrong to say that social media network is playing a major role in shaping up our behavior and thoughts now-a-days. The practice of instant thought or opinion sharing to a huge number of individuals on social media has made internet Polarized which means that whenever a certain incident happens it always divides the people on the basis of what they believe and just like our daily life, users having different types of personalities participate in such online trends and campaigns. Twitter is the famous example of social media where online mobs use trending hash tags to show their opinions and thoughts related to a specific incident. The purpose of this study is to investigate what kind of users participates in such mobs by analyzing their network features and compare two kind of mobs to analyze how different they are from each other in terms of those features by using different techniques of Data Science and Natural Language Processing. In the end, users/participants of both mobs/campaigns, with respect to their tweets, will be analyzed personality wise to classify them as “Normal user” or “Aggressive user”.