Home▸Data Science▸Link Prediction Lab: Node Embeddings & Edge Scoring

Link Prediction Lab: Node Embeddings & Edge Scoring

Interactive 3D graph embedding simulator: watch a GCN-style neighbor-aggregation rule pull node embeddings together, then score held-out links by cosine similarity or dot product and watch the predicted-vs-missed links light up with a live AUC readout.

Data Science3DAdvanced60 FPS

This simulator builds a random community-structured graph, gives every node a random 3D embedding, and hides a fraction of the real edges as a test set. A GCN-style neighbor-mean aggregation step — the same mechanism behind GCN, GraphSAGE and GAT-style message passing — repeatedly pulls each node's embedding toward the average of its visible (train) neighbors. Score every non-train pair by cosine similarity or raw dot product, sweep a threshold to see which held-out links get correctly predicted versus missed, watch false-positive predictions light up, and track a live AUC estimate as the embeddings converge.

⚙ Under the hood

Watch a GCN-style neighbor-mean aggregation pull node embeddings together on a live 3D graph, then score held-out edges by cosine similarity or dot product and track a live AUC as predicted, missed, and false-positive links light up.

graph neural networkslink predictionnode embeddingsGCNdata scienceAUC

3D · Three.js / WebGL renderer · 60 FPS target · runs fully client-side, no install