Loading
205 articles, 2497 links between them. Node size is degree — how many other articles a topic touches. Colour is the primary portal by default, and can be switched to evidence grade, readiness, or reversibility. Drag a node to pull the layout apart, scroll to zoom, click a node to isolate its neighbourhood, and pick two to trace the shortest path between them.
The graph is built from wikilinks, not from tags or categories. If one article references another in its prose, an edge exists. That makes the shape an artefact of how the subject actually hangs together rather than of how it has been filed.
A few things fall out of it. The longevity cluster is the densest, which reflects both the volume of work in that area and the fact that its concepts genuinely depend on one another. Articles on the periphery are usually organizations and people — they cite the science, but the science does not need to cite them back.
Reciprocal edges, drawn slightly heavier, mark the 990pairs where both articles reference each other. Those are the strongest signal of genuine adjacency, and they are what the “related articles” ranking weights most.
The most-referenced articles. These are the concepts the rest of the corpus leans on; if you are new to the subject, they are the right place to start.