Name Map

What You're Seeing

Each dot is a name. The colours represent the 10 clusters that a grouping algorithm (K-Means) placed the names into. Names that sit close together on the map are similar in sound and era. If they share a colour, they belong to the same cluster.

How the Map Was Made

Each name is described by around 55 measurements. 55 dimensions are impossible to plot on a flat screen, so a technique called t-SNE (t-Distributed Stochastic Neighbour Embedding) is used to squash all the information down to two dimensions. It tries to keep similar names near each other and push different names apart, but some information is always lost in the squash.

The 2D Limitation

Two dots that appear far apart here might actually be quite close in the full multi-dimensional calculation, and vice versa. Use the map as a rough visual guide; for exact distances between names, the Name Neighbours page is more precise.

Navigation

Scroll to zoom in, click and drag to pan, and type in the search box to jump to a specific name. Click any dot to hide it. Use the sidebar to toggle entire clusters on or off.