--- import Layout from '../layouts/Layout.astro'; import { SemanticVisualization } from '../components/react/SemanticVisualization'; ---

Nouns Semantic Visualization

Explore Nouns NFTs based on their semantic similarity using vector embeddings with exponentially adjusted spacing. Similar nouns are pushed slightly apart while distant nouns are pulled closer together for better visualization.

How it works

Each Noun is represented by a vector embedding capturing its visual features. We calculate cosine similarity and apply an exponential distance adjustment: close nodes (high similarity) are pushed apart using e^(2x), while distant nodes (low similarity) are pulled together using e^(-x), creating a more balanced layout that prevents overcrowding of similar items.

Interaction Guide

  • Click a noun to highlight its similarity connections
  • Drag nodes to reposition them
  • Zoom in/out and pan around the visualization
  • • Adjust similarity threshold to show more or fewer connections
  • • Change max nouns to include more data in the visualization
  • Colors represent different noun categories