Is it Viral?
Finding the secret recipe for a viral video is extremely enticing as many of the popular YouTube celebrities earn millions off of advertising revenue. In a paper from Chapman University, researchers attempt to derive a mathematical calculation for the probability that a piece of content goes viral. Using Facebook networks as a research tool, the group is attempting to find the best methods to diffuse information across Facebook networks. In this case, the focus is on health related information, but the findings can extend to a wide variety of content.
In class, we have talked about information sharing across a network and also briefly covered the SIR model of epidemics. The spread of viral content can roughly be modeled by a similar SIR network where each node has either not seen the content, seen the content and not shared it, or seen the content and shared it. Based on our approximation of the reproductive number in this network, we can roughly estimate if some content will continue spreading in the network or die out. It is, however, extremely hard to predict the fraction x, or how likely a person is to share the content. The Chapman research group hopes to use Facebook to derive a mathematical expression as they believe it to be an excellent place to test information spread in a network as it is plays a central role in many people’s daily life. By identifying Facebook friends and users who share the same behaviors, likes, and groups, they are able to better estimate how likely a certain demographic is likely to spread content.
The focus on health information is important as it will lead to optimal ways to help spread critical information. The research groups plans to run simulations to determine the optimal way to begin to spread information throughout the network without it dying out after a few initial waves.
Source: http://www.sciencedaily.com/releases/2015/10/151006132031.htm