Structure in scientific networks: towards predictions of research dynamism

TL;DR

This study investigates whether structural patterns in citation networks can predict the growth or decline of scientific research areas, using optical physics as a case study and applying network analysis techniques.

Contribution

It introduces a method combining weak tie theory and entropy measures to analyze citation network structures and their relation to research area dynamics.

Findings

Citation network structures reflect research area trajectories.

Networks of expanding areas show different patterns than declining ones.

Structural features may influence scholarly conversations and research evolution.

Abstract

Certain areas of scientific research flourish while others lose advocates and attention. We are interested in whether structural patterns within citation networks correspond to the growth or decline of the research areas to which those networks belong. We focus on three topic areas within optical physics as a set of cases; those areas have developed along different trajectories: one continues to expand rapidly; another is on the wane after an earlier peak; the final area has re-emerged after a short waning period. These three areas have substantial overlaps in the types of equipment they use and general methodology; at the same time, their citation networks are largely independent of each other. For each of our three areas, we map the citation networks of the top-100 most-cited papers, published pre-1999. In order to quantify the structures of the selected articles' citation networks, we use a modified version of weak tie theory in tandem with entropy measures. Although the fortunes of a given research area are most obviously the result of accumulated innovations and impasses, our preliminary study provides evidence that these citation networks' emergent structures reflect those developments and may shape evolving conversations in the scholarly literature.