Complex network analysis of keywords co-occurrence in the recent efficiency analysis literature

• Published in: Scientometrics Vol. 120; no. 2; pp. 609 - 629
• Main Authors: Lozano, S., Calzada-Infante, L., Adenso-Díaz, B., García, S.
• Format: Journal Article
• Language: English
• Published: Cham Springer International Publishing 15.08.2019; Springer Nature B.V
• Subjects: Community structure; Computer Science; Data envelopment analysis; Information Storage and Retrieval; Keywords; Library Science; Network analysis; Operations research; Complex network analysis; DEA; Emerging topics; Keywords; Co-occurrence network; Research gaps
• ISSN: 0138-9130; 1588-2861
• DOI: 10.1007/s11192-019-03132-w

A network analysis (NA) of keyword co-occurrences for a broad collection of Data envelopment analysis (DEA) papers in the period 2008–2017 is carried out. The raw keywords have been cleaned up and standardized to consolidate and increase the consistency of the keywords. The resulting network has been characterized using network-level as well as node-level NA measures. Although the size of the network steadily increases with time, the average path length does not, showing its small world character. The disassortativity of the network indicates that the keywords used in a given paper generally include one or more common, frequently-used terms plus other less common terms that refer to the specific context of the research. The evolving nature of the keyword network is highlighted with some DEA keywords staying at the top of the ranking during the whole period and other emerging topics significantly increasing their strength during this period. The community structure of the network, which reflects the field’s knowledge structure, is also presented. The identified communities generally include specific DEA methodology terms, linked with corresponding application areas as well as with some geographical references. Also, the ego-network of some sample keywords is shown, and some research gaps in DEA are identified.