Beyond affiliation: Geographic and topical visualization of the world ocean assessment

The United Nations World Ocean Assessment II (WOA II) synthesizes oceanographic research through 49 chapters citing over 3500 scientific articles. Exploring a corpus of this scale requires users to connect its document structure with topical, temporal, citation, and geographic information. Existing bibliographic metadata can describe where researchers are affiliated, but does not necessarily represent the geographic locations discussed in the research itself. We present an interactive scholarly data visualization system whose central contribution is the integration of content-derived geographic information with WOA II chapter structure and OpenAlex topical classifications in a coordinated exploration environment. Named entity recognition is applied to article abstracts to identify geographic references, a subset of which are matched against a controlled vocabulary representing 12 canonical marine regions and incorporated into article-level and corpus-level views. A three-column relationship view supports exploration from either a geographic or document-centered starting point, while coordinated topic, temporal, article, and reference views support further inspection of the associated literature. A case study demonstrates how the system can be used to move from geographic and topical relationships to supporting articles and references. Although developed for WOA II, the approach is applicable to other domain-specific assessment corpora in which geographic information can be extracted from document content.