Message Propagation and Social Influence in Twitter

Abstract

Twitter data has potentially unlimited value and numerous applications and is known for its increase in users over time. Twitter facilitates information diffusion at an exponential rate and also the creation of networks of users with a common interest. People reacting to the spread of an epidemic or a natural disaster are greatly influenced by the information diffusion in social media. Twitter, being a popular micro-blogging network provides an effective way to measure diffusion in terms of speed and strength. Our research is based on previous work on models related to topic diffusion and user influence. A topic is defined by a set of keywords.This research concentrates on the implementation of algorithms for computation of diffusion of a topic in twitter. The degree of influence of the users who tweet on the topic is also addressed. We have presented two different approaches to compute user influence based on topic potential. We compare two diffusion models proposed in the literature, namely potentials and connections. For testing and empirical analyses we use tweets related to “flu”, “food poisoning”, and “politics”.