3D/2D Text Visualisation

Like everyone, I use 3js for data visualisation. But while semantic clusters are interesting, they don't confer much practical information alone.

So I did something very simple: a query spatially re-assembles the nodes, and you can switch them to text.

Honestly, it's hard to swing 3D viz for text as a genuinely useful feature and not a novelty, but I get the feeling there's still some potential in the idea. Would be good to discuss, I'm sure someone here has made a better implementation.