Study on Influencers of Cryptocurrency Follow-Network on GitHub

Abstract

Open-source software (OSS) is widely used and has become an essential infrastructure for our society today. Substantial research has been done to improve the success of OSS development. Of them, studies about influencers have gained attention in recent times. Influencers are regarded as an evangelist in a specific domain, and play an important role in persuading others. They are frequently analyzed on Twitter and other SNSs. With the advent of social coding platforms such as GitHub, research has started on OSS influencers who seem to affect the behavior of developers. However, there is not yet enough research on the method of identifying influencers and their effects on OSS. In this study, we analyzed the follow-network of cryptocurrency projects developed on GitHub quantitatively, and found (1) The HITS algorithm is more effective when compared with in-degree centrality and PageRank algorithm in identifying influencers of a specific domain. (2) The rate of contribution of a user correlates with their rate of influence, but the explanatory power is small. The amount of activity on GitHub is not as essential for OSS influencers as it is on Twitter, which requires a lot of activity to be an influencer. (3) The rate of influence of influencers on a project correlates with the number of contributors.