Tutorial — Network mapping quickstart
Last updated: October 21, 2026
Skill: network-analysis-specialist
Time: ~40 minutes
Goal
Turn a messy list of entities and relationships into a small inspectable graph.
Steps
- Export nodes (people, companies, URLs) and edges (owns, paid, employed, cited) to CSV.
- Import into Gephi or Graphistry / org-approved tool.
- Apply force-directed layout; label only high-degree nodes to reduce clutter.
- Compute degree; flag hubs and bridges for manual review.
- Color by entity type; verify no accidental PII in node labels.
- Export PNG + save project file for reproducibility.
Choosing the tool and reading the layout
Gephi, named in step 2, is free, open-source desktop software that handles networks of up to about one million nodes and five million edges. Besides drawing layouts it computes the measures step 4 needs — degree, betweenness centrality, closeness, clustering coefficient and PageRank — and finds communities by modularity. Its dynamic filters cut a network down to the nodes and edges that answer one question, which is the practical fix for the hairball in Pitfalls. Its native GEXF format is also an export option in OSINT tools such as SpiderFoot, so scan results can be opened directly.
Step 3’s force-directed layout is, in Gephi, usually ForceAtlas2, published by Mathieu Jacomy and colleagues in PLOS ONE on June 10, 2014. It simulates a physical system: connected nodes attract each other, every node repels the others with a force scaled by its degree, and the drawing settles where the forces balance. The authors state the limit plainly — the position of a node cannot be interpreted on its own; it has to be compared to the others. Nodes drawn close together tend to share connections. Nearness on the screen says nothing about who controls, pays or employs whom.
So report the measure, not the picture. “Highest betweenness of the 140 entities” is a finding a reader can check against your exported project file; “sits at the centre of the map” is an artefact of the layout. A node with high betweenness lies on many of the shortest paths between other nodes — in a donor or ownership network, often an intermediary worth a records request.
Pitfalls
- Hairball graphs convince nobody — filter to a question.
- Layout implies nothing about causality.
Next steps
- Tutorial — Beneficial ownership tracing for corporate chains
Sources
- Gephi, Gephi Desktop features — scale, statistics, filtering.
- Jacomy M, Venturini T, Heymann S, Bastian M (2014), ForceAtlas2, a Continuous Graph Layout Algorithm for Handy Network Visualization Designed for the Gephi Software, PLOS ONE 9(6): e98679.
- SpiderFoot, project repository — CSV/JSON/GEXF export.
Expanded with AI research and drafting assistance from the primary sources cited; editorial review pending.
