An Exploratory Data Analysis of Fake News on Twitter From the Perspective of Pakistani Tweets

Abstract

Nowadays, social media has become the most popular way of communication among the people. People share everything on social media, even things that they do not know the authenticity of, or have no knowledge about. In Pakistan, typically the people just share all sorts of news without verifying it. They don’t bother with ensuring whether the news is true or not and just share things for the sake of gaining likes and shares. Due to extensive use of social media, and careless sharing of news, an important issue is fake news. Whenever a new event or new issue arises in Pakistan, the people started posting fake news regarding that event or issue in hopes of getting lots of likes and views. One aim of such sharing of unverified news is that many people earn money from this as well. If we talk about politics in Pakistan, fake news is often employed as a tool by the opposing parties to discredit each other at the start of elections. There many examples of fake news in foreign countries and a lot of research have been done there for the sake of identifying these fake articles. This thesis wishes to explore the fake news characteristics due to which they got diffused in Pakistan using social media platform “Twitter”. The study performs statistical analysis and uses data mining techniques to find out the characteristics of fake news or tweets and explores the relationship between different features of tweets and presents its findings so that they can be used for the future research in Pakistan on fake news and its implications.