Searching for temporal patterns in the time series of publications of authors in a research specialty

Abstract

In this paper we report results of our investigation of temporal patterns in the publication activity of authors in a research specialty. We base our analysis on Web of Science data for a field in the physical and chemical sciences from 1991-2012. We determine the research groups in the field by clustering the co-author network and generate our sample for this analysis by selecting the most productive author of each co-author cluster to represent the activity of that group. Whereas a statistical time series analysis did not reveal any specific patterns, a time series clustering approach generated a grouping of time series that correlates with the structural network position (‘node role') of the respective authors in the clustered co-author network. This work is part of a long-term research project employing a mix of qualitative and network analytic methods to investigate field-specific differences in collaborative patterns.