A complement to the novel disruption indicator based on knowledge entities

Abstract

Following the proposal of disruption index (DI) for detecting scientific breakthroughs based on citation patterns, a recently introduced knowledge entity-based disruption (ED) index incorporates both citation patterns and knowledge elements. In this study, we investigate the applications and limitations of the ED series indicators by employing two datasets from different fields within the Web of Science database, providing some insights that complement the use of ED series indicators. For the genome editing dataset, we validate the consistency across the ED series indicators based on different knowledge entities, specifically MeSH terms and KeyWords Plus. In the case of the h-set dataset, where no MeSH terms were matched, our focus is on comparing the performance of the ED series indicators based on KeyWords Plus with other representative disruption indicators in small datasets. When considering the two datasets of the “stem” and “seed” papers obtained by the seed algorithm as reference objects and calculating their DI and ED series indicators, the results indicate that the values of DI series indicators of “seed” papers exhibit higher values compared to the ED series indicators. From a statistics perspective, there are no significant differences in the ED series indicators when employing different knowledge entities, despite variations in their rankings. Based on the results and discussions of this study, we provide guidance on application of ED series indicators and potential refinements in subsequent studies.