Contents

VosviewerEdit

VOSviewer is a software tool designed for the construction and visualization of bibliometric networks. It helps researchers map relationships among authors, journals, and terms, revealing collaboration patterns, topical clusters, and the diffusion of ideas across disciplines. The program was developed by researchers at Leiden University's Centre for Science and Technology Studies (CWTS) and by the engineers Marc van Eck and Ludo Waltman. Since its first release around 2010, it has become a staple in science mapping and bibliometrics, praised for making large bodies of scholarly data approachable and interpretable.

VOSviewer supports data from major bibliographic databases such as Scopus, Web of Science, and Dimensions as well as PubMed, enabling users to build maps from diverse sources. Its outputs are designed to be interoperable with common graph and visualization workflows, including exporting to formats used by Pajek and Gephi and producing publication-ready visuals. Emphasis is placed on transparency and reproducibility: maps can be recreated from the raw data, and researchers can adjust parameters to test the robustness of findings.

Overview

  • Purpose and scope: VOSviewer focuses on the visualization of bibliometric relationships to help scholars understand current research landscapes, collaboration networks, and topic development across time.
  • Visualization modes: It offers network maps, density (heat) maps, and overlay visualizations that can highlight changes over time or by another dimension such as author affiliation or journal category.
  • Core relations canvassed: Co-authorship networks, co-citation networks, bibliographic coupling, and term co-occurrence maps are among the primary graph types produced.
  • Output and interoperability: Users can export high-quality images for publication and data files for import into other tools, including common graph formats and the VOSviewer-native data format.

History

VOSviewer originated from work at CWTS at Leiden University, with the initial goal of enabling researchers to visualize large-scale bibliometric data in a user-friendly way. The project was developed by Marc van Eck and Ludo Waltman and quickly gained traction in libraries, research centers, and university departments looking for an objective, repeatable way to explore scholarly communication. Over the years, the software expanded to support overlay visualizations, improved scalability, and broader data import options, while maintaining a clear emphasis on reproducible mappings of scientific activity.

Technical foundations

  • Visualization of Similarities (VOS): The core mapping technique places items in a two-dimensional space so that proximity reflects similarity derived from the chosen relationship (co-authorship, co-citation, etc.). This approach yields intuitive maps in which related items cluster together.
  • Similarity measures and network construction: Depending on the chosen relation, VOSviewer computes similarity scores (e.g., co-authorship strength, citation links, or term co-occurrence) to build the underlying network.
  • Clustering and interpretation: A clustering step groups nearby items into communities, which are then color-coded to reveal thematic or disciplinary clusters. The clustering approach draws on community-detection techniques commonly used in network science, such as variants of the Louvain method.
  • Data handling and visualization: The software emphasizes legible, scalable visuals and provides options to filter networks by size, density, or time, helping users focus on the most informative portions of a map.

Data sources and limitations

  • Primary data sources: VOSviewer can import data from major bibliographic indexers such as Scopus, Web of Science, and Dimensions, as well as domain-specific databases like PubMed. The choice of data source shapes the map, given differences in coverage, indexing practices, and author name variants.
  • Data quality considerations: Name disambiguation, affiliation changes, and incomplete metadata can affect the accuracy of co-authorship and institution-related analyses. Users should be mindful of these limitations and, when possible, corroborate findings with multiple data sources.
  • Scope and bias: Bibliometric maps reflect publication and citation patterns, which can be influenced by disciplinary norms, language coverage, funding ecosystems, and the indexing biases of chosen databases. Readers should interpret clusters and connections as representations of scholarly activity, not as definitive judgments about quality.

Applications

  • Literature surveys and mapping: Researchers use VOSviewer to survey large literatures, identify key topics, and visualize how fields evolve over time.
  • Evaluation and planning: Universities and research centers employ maps to understand collaboration patterns, identify established or emerging research programs, and guide strategic decisions.
  • Policy and portfolio analysis: Funding bodies and decision-makers can use bibliometric visualizations to monitor research activity across disciplines and regions, informing resource allocation.

Controversies and debates

  • Interpretation and robustness: As with any visualization-based method, maps depend on parameter choices (e.g., similarity metrics, thresholds, and clustering resolution). Critics argue that different settings can yield different structures, which can lead to over- or under-emphasizing certain relationships. Proponents counter that transparent reporting of parameters and replication across data sources mitigate these concerns.
  • Metrics and merit, not ideology: Bibliometric tools are sometimes viewed through the lens of broader debates about how scholarly impact should be measured. Supporters argue that objective, data-driven indicators support efficient allocation of research resources and accountability, while critics contend that overreliance on metrics can distort incentives and overlook qualitative contributions. From the perspective of those who favor practical optimization of research ecosystems, transparent metrics and defensible mappings are tools to improve outcomes rather than obstacles to progress.
  • Inclusivity and coverage: Critics sometimes claim that bibliometric analyses inadequately capture non-English literature, regional publishing, or early-career researchers. The response from practitioners is that expanding data sources and improving author disambiguation can address gaps, while still valuing the clarity and comparability that standardized maps provide. Proponents emphasize that maps are best used as one part of a balanced evaluation framework, not as the sole determinant of scientific merit.

See also