Modeling Rumors in Twitter: An Overview

1. Introduction

Social networks are very powerful means of communication where information can flow fast and has deep penetration. Any user can both generate and consume content which is provided to wider audience when compared to conventional media. Social media has power to affect user behavior and emotions as shown by Robert M. Bond et al. (2012) in an experiment conducted over 61 million subjects, where they found that 2% more users have voted when associated with friends who have voted and shared on Facebook. Facebook Scientist Adam Kramer et al. (2014) has shown direct impact of a Facebook post on the emotion of a user. It also showed flow of an emotion over the social networks. Engaging topics like politics, religion, race and etc. have even higher effect on the users. Egyptian revolution is one such case, which distinctly depicts the impact of Facebook (Bradly, 2008). Video of Khaled Said an Egyptian businessman who was beaten to death by police in June 2010 was leaked on YouTube and a page named ‘We are all Khaled Said’ was created on Facebook to protest. The page was joined by hundred thousand citizens and played prominent role in spreading the discontent among public. The page called for protest on 25^th January, which witnessed 400,000 participants. The revolution ended with resignation of the president. Trends on social networks are extracted using Data Mining algorithms. The study, which extracts meaningful content from large datasets, is called data mining. Data mining has its roots penetrated to different segregated sectors like data mining is used with Rough Set Theory to extract meaningful knowledge from large databases (Rana & Lal, 2016), data mining is used to mine medical data (Banu et al., 2015; Dey et al., 2014), it can be used with steganography techniques (Bhattacharya et al., 2012) and it is also used in systems proposing frameworks for firms dealing customer management relationship (Ranjan & Bhatnagar, 2009).