A systematic identification and analysis of scientists on Twitter
Top: Distribution of in-degree d← or in-strength s←, PageRank (PR), and k-core number k in the three networks. Middle: Portion of centralities occupied by scientists in each group. We calculate, for each centrality in each network, the sum of centrality values of users in each scientist group divided by the total centrality values. Bottom: Normalized portion of the three types of centralities occupied by scientists in each group in the three networks. Normalization is done by dividing the portion by number of scientists.