A live map of the news cycle, rebuilt hourly from embedded headlines

Name: A live map of the news cycle, rebuilt hourly from embedded headlines
Availability: InStock
Author: sach90

by sach90·Jul 23, 2026·4 points·3 comments

Visit Project View on HN

AI Analysis

●●●BangerRabbit HoleBig BrainEye Candy

Recursive DBSCAN clustering turns 150k chaotic headlines into a navigable galaxy of news stories.

Strengths

•Full client-side interactivity after initial 18MB load means zero latency panning or zooming.
•UMAP layout plus recursive DBSCAN creates genuinely coherent topic regions, not just random scatter.
•Hourly full rebuild from scratch ensures the map reflects the live news cycle, not stale caches.

Weaknesses

•18MB initial payload is a hard barrier for mobile users on cellular networks.
•Lacks temporal playback; seeing how clusters evolve over days would be more valuable than static hourly snapshots.

Post Description

Hi HN. This is a live map of the English-language news cycle, with headlines are ingested continuously from Google News, US and UK editions.

Each one is embedded with OpenAI's text-embedding-3-large, clustered into stories with hnswlib, laid out in 2D with UMAP so related topics sit near each other, grouped into topic regions with recursive DBSCAN, and labeled by an LLM. The map is rebuilt from scratch every hour.

I deliberately skipped progressive loading. The whole map loads upfront, ~18MB of data, and after that everything runs client-side with zero network requests: pan, zoom, search, cluster panels. deck.gl handles all 133k points fine, even on mobile.

Currently showing ~133,000 stories across 1,081 topics.

Happy to answer any questions.