Mining the Most Influential Authors in Academic Publication Networks through Scholastic Actions Propagation

The influence degree of academic authors can be judged in one way by citation-based indices such as the h-index [1]. Marketing researchers develop a theory of action propagation to measure consumers’ influence in a market; by going viral, an action of marketing essential such as product adoption or brand awareness is spread to the majority of a market autonomously. Selecting the most influential consumers as marketing seeds is a hot topic in marketing research. In this study, we mine the most influential authors from the perspective of their ability to propagate scholastic actions such as attending a serial conference or publishing in a specific journal. The credit distribution model from Goyal et al. [2] is chosen as the propagation model with two academic publication networks (citation and coauthoring). Real data consisting of 10 years publication records from DBLP and ACM were used in the experiments. It is found that the citation network is more efficient than the coauthoring network to propagate scholastic actions. Top influential authors who can effectively affect fellows to attend a serial conference or publish in a specific journal are mined with the citation publication network.

.