Exploratory experiment on co-authorship network using social network analysis metrics and measures

Abstract

This paper contributes in understanding and gaining meaningful insight about the relationship among the scientist in the co-authorship network using social network analysis. We argue that the relationship analysis is not always a straightforward process. In the past one single measure, for example, the egocentric or centrality measure was used to describe the scientific collaboration patterns separately. In this paper, various analysis such as centrality analysis, ego network, community detection, largest clique and word frequency have been used to examine and interpret the collaboration among the authors. This research is not dominated by known researchers but involves an overall exploration of the network. Our research is mainly guided by the creation of research issues, assessing the type of dataset and the objectives for presenting the co-authorship relationships. It is important to identify the motive of the selected measures in order to achieve the predefined objective. Specific methodology and procedures are designed to solve each research issue respectively. This study reveals that the network interpretation should not be solely based on one network measure, but an explorative analysis results need to be considered because it allows exploring the hidden information through the changes in the network structure, topology patterns and nodes' position.