Exploring distributional characteristics and similarities of scholarly keywords: a comparative study of Web of Science Keywords Plus and Dimensions Concepts

Purpose The goal of this study is to assess the degree of resemblance between machine-generated terms provided by two major indexing systems: Web of Science Keywords Plus and Dimensions Concepts. Design/methodology/approach A thorough analysis examines the distributional characteristics and similarities between these two terms. The study utilizes the rank frequency distribution of terms and comparisons of their forms using goodness-of-fit measures to assess distributional properties. Whereas to evaluate the similarities, the study utilized Jaccard similarity measures between high-frequency terms as well as overall terms (i.e. KW Plus and Dimensions Concepts). Findings The findings demonstrate that these two terms differ significantly in both distributional forms and similarities, thus representing different kinds of information related to the publication. The findings further indicate that the algorithms used by both databases for term generation/extraction are quite different from each other. Research limitations/implications The implications of this study will enhance scholarly indexing and retrieval practices, supporting effective information access, organization and interdisciplinary research within academic databases and knowledge systems. Originality/value The novelty of the study is that it focuses on revealing the characteristics, similarities and differences between major indexing terms that were previously argued to be useful for performing various text analysis and scientometric exercises.