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
