A Data Science Expert Social Network: From Personal Follower List to Social Network Structure

Abstract

The researchers combined two lists of 28 favorite “data science experts to follow on Twitter” to seed a Twitter network and analyze whether the recommended experts were indeed amongst the most influential “data science experts” on Twitter. They analyzed the resulting Twitter network to find the most important nodes in terms of popularity, quality of connections, types of roles played, such as bridges, and node ability to quickly spread information. They found that only some of the recommended experts appeared most influential given the network analysis. They also found that the experts on the list landed mainly in two sub-groups. Starting with a writer's favorite list of experts may be helpful in seeding a more comprehensive list.