A method for mapping bibliometric relations based on field-classifications and citations of articles

Abstract

We introduce quasi-correspondence analysis, a new method for mapping interrelations between publishing scientific entities, e.g. journals. It is based on the spatial configurations resulting from the analysis of the following bibliometric data: the relations in the citation-structure, which are examined for both the cited and citing-mode of scientific entities, and the distributions of the frequencies of different field-classification codes assigned to the documents published by each entity. The network of relations are visualized ('mapped') as separate configurations of points in a low-dimensional space. Each point represents a row or column of the corresponding data matrices. Distances between points represent the strenght of the relationship between scientific entities with respect to field-classifications of documents and citation-flows, respectively. Both configurations are then combined into a joint interpretation of the relations between the entities.