The message does not matter: the influence of the network on information diffusion (1508.05617v1)

Abstract: How an information spreads throughout a social network is a valuable knowledge sought by many groups such as marketing enterprises and political parties. If they can somehow predict the impact of a given message or manipulate it in order to amplify how long it will spread, it would give them a huge advantage over their competitors. Intuitively, it is expected that two factors contribute to make an information becoming viral: how influential the person who spreads is inside its network and the content of the message. The former should have a more important role, since people will not just blindly share any content, or will they? In this work it is found that the degree of a node alone is capable of accurately predicting how many followers of the seed user will spread the information through a simple linear regression. The analysis was performed with five different messages from Twitter network that was shared with different degrees along the users. The results show evidences that no matter the content, the number of affected neighbors is predictable. The role of the content of the messages of a user is likely to influence the network formation and the path the message will follow through the network.