Cocitation

Co-citation is a concept in bibliometrics describing the frequency with which two documents are cited together by later works. When a subsequent paper cites both documents A and B, A and B are said to be co-cited; the strength of their co-citation is the number of times they appear together in citations. Co-citation analysis uses these frequencies to identify related literature and to map the structure and evolution of scientific fields.

Co-citation differs from bibliographic coupling. Co-citation is based on how earlier works are cited together by later papers, whereas bibliographic coupling links documents that share common references in the same document. Co-citation can be conducted at multiple levels, including document co-citation (two papers), author co-citation (two authors cited together), and journal co-citation (two journals cited together).

The method originated with Henry Small in 1973, who introduced co-citation analysis to study the structure of science. Contemporary approaches construct a co-citation matrix, apply normalization (such as cosine similarity), and use network analysis and clustering to reveal citation communities, research fronts, and influential works. Visualization and measures of centrality, modularity, and stability aid in interpreting the evolving landscape of a field.

Applications include mapping scientific domains, tracking topic development over time, identifying foundational papers, informing literature reviews, and supporting information retrieval and recommendation systems. Data for co-citation analyses are typically drawn from bibliographic databases like Web of Science or Scopus, by extracting cited references. Limitations include dependence on citation practices, time lag, disciplinary differences, and potential biases toward older or highly cited works; careful normalization and interpretation are required.