VOSviewer

VOSviewer is a free tool for constructing and visualizing bibliometric networks, developed by Nees Jan van Eck and Ludo Waltman at the Centre for Science and Technology Studies (CWTS), Leiden University. The original application is a Java desktop program that builds co-authorship, co-citation, bibliographic-coupling, and term co-occurrence maps from bibliographic data (Web of Science, Scopus, PubMed, and similar exports). Node positions are computed with the VOS ("visualization of similarities") layout technique, and items are grouped into clusters shown as a colored map. Although designed for scientometrics, VOSviewer is not restricted to bibliometric data and can render any network supplied in its file formats.

VOSviewer-Online is a web-based reimplementation of the viewer, built in JavaScript with React, Material-UI, and D3 by the same CWTS team with Olya Stukova and Nikita Rokotyan. It is distributed as a hosted app (app.vosviewer.com) and as the npm React component vosviewer-online, embeddable via iframe or as <VOSviewerOnline data={...} parameters={...} />. Unlike the desktop program it only visualizes precomputed maps; it does not perform the counting, clustering, or layout steps itself.

The data model uses two legacy tab-delimited files or a single JSON file. A map file is one row per item with a header line and columns such as id, label, x, y, cluster, plus repeatable weight<...> and score<...> columns (weights size the nodes; scores drive the color overlay). A network file lists links as two-item-ID pairs with an optional third strength column. The consolidated VOSviewer JSON format nests a network object (arrays of items with id, label, x, y, cluster, weights, scores, url, description, and links with sourceid, targetid, strength) alongside config and info sections that carry colors, terminology, and metadata.

Because it uses this idiosyncratic schema rather than a general interchange format, VOSviewer sits at the edge of the VOSviewer conversion ecosystem: exporters and scripts translate between it and mainstream formats such as GraphML, GML, and Pajek, which VOSviewer's desktop version can also read. It overlaps in purpose with general network viewers like Gephi and Cytoscape but is more specialized and opinionated. Its strengths are polished, publication-ready thematic maps, density and overlay views, and effortless web embedding; its main limitations are the requirement that coordinates and clusters be precomputed (the online viewer does no layout), weak support for directed or attributed graphs beyond weights and scores, and a niche file format that needs conversion for interoperability.