Semantic || Spatial Networks

Map View:

This project visualizes the spatial and semantic similarity of blocks in Manhattan. I generated two sets of embeddings, from physical features [building density, height, lot area, year built, etc] and semantic data [place types, names, reviews, etc]. I then clustered the block embeddings with k-means and visualized each cluster as a network layer. The resulting interactive web map enables users to explore and compare the semantic and spatial topology of Manhattan.