Influence Cascades and Behavioral Influence Patterns in Social Media
An article published through MDPI describes the effects of influence cascades and how they are constructed as a result of social media influence. This article gives an overall approach of different OSN’s and gives analysis that specifies how these different OSN’s (twitter and GitHub) are significantly effected by social media influence regarding cryptocurrencies in addition to CVE’s. Influence cascades are defined as every single action within a chain that starts with a preliminary user who was effected by an external stimulus then that preliminary user will proceed to influence other individuals within that chain. This occurs within the chain until there is no user that influences any other individual within the chain. In the research done it is such that the individuals are only looking at users that are not influenced but rather influence others within the OSN system. The two social media platforms discussed in the article are Github and Twitter, we will notice how the researchers evaluate how Twitter and Github users discuss cryptocurrencies.. The system that the researchers use to look at an influence cascades is such that there is a user they refer to as u who performs an action A such that her action will effect another user v in a way that is denoted as b. Looking at a system of two levels of influence such that total social influence from user u to v as a vector −→γ u,v would be reflected by the diagram below. Keep in mind each node will be disclosed as u in the following diagram because we consider these nodes from the influencer perspective as the influence is continued. 
The research from the paper specifically looks at twitter crypto social influence data by extrapolating data from differing websites for more than 20 “target coins” and connecting this data with references to target coin names in addition to codes and hashtags. In addition, for Github crypto currency data was extrapolated from events related to these coins’ information disclosures in addition to information keepers for target coin names and descriptions. In essence the researchers connected certain topics with their discussion and prevalence on twitter and Github. The researchers then created a scale free network of 1406 nodes where they evaluated the root nodes (origins of the cascade) and situated all the influence cascades by level and evaluated normalized total influence in the same way that is disclosed in the graphic above. The researchers then found statistical results that concluded that cryptocurrency disseminated through influence cascades at a rate that is much shorter than expected compared to scale free null-models which serves as a control to the empirical data. These results are substantiated by looking at average user’s distributions over influence cascades for the crypto community being relatively robust across the twitter and GitHub platforms. In addition the findings disclosed that for Github cryptocurrency and Twitter cryptocurrency information there is shared result for their influence cascades such that the highest fraction of the influence flow occurs during the middle of the cascade. This could mean that when individuals disclose information it does not gain an impact until everyone thinks that everyone else has adopted this belief then the steam of this belief begins to tail off towards the end of the cascade.
This research regarding influence cascades for information on cryptocurrency and its importance within the platforms of Github and Twitter holds great relevance within the tech savvy age that is currently overtaking the global economy. In addition it is very important to understand how technology, in addition to the impact of networks, changes peoples perception of how they will act in reference to how other people act.