license: cc-by-4.0
pretty_name: Dead-Web Common Crawl (2018-2026)
tags:
- web
- link-rot
- common-crawl
- dead-domains
- reachability
configs:
- config_name: sample
data_files: data/sample.jsonl
Dead-Web Common Crawl — a longitudinal host-reachability panel (2018-2026)
An open dataset labeling the reachability of 172,959,928 registered domains across 80 monthly Common Crawl archives (2018-2026), built from Common Crawl's robotstxt subset and calibrated against a live re-probe to separate genuinely-dead domains from those merely blocking crawlers or dropped from the crawl.
40M domains (23% of the 8-year union) are calibrated genuinely dead. Of domains that disappeared
from Common Crawl, only 64.2% [63.5-64.9%] are truly dead — 31% still resolve, 4.7% are blocked.
Domains last seen blocked are only 44.7% dead (the dead-vs-blocked distinction, measured).
Per-domain schema: domain, first_ord, last_ord, total_crawls(80), n_present, last_state, recent_absence_streak, label (alive_present/dead_candidate/dark_ambiguous/intermittent).
Built with the open-source crawlora-deadweb CLI (https://github.com/Crawlora-org/crawlora-deadweb). Source repo: https://github.com/Crawlora-org/dead-web-commoncrawl . License: CC BY 4.0.
Citation
If you use this dataset, please cite the Zenodo archive — concept DOI
10.5281/zenodo.20814397 (always resolves to the latest version; this release, v1.0.0,
is 10.5281/zenodo.20814398):
Crawlora (2026). Dead-Web Common Crawl: a longitudinal host-reachability panel (2018-2026) [Data set]. Zenodo. https://doi.org/10.5281/zenodo.20814397
Mirrors: Zenodo (DOI) · GitHub · Kaggle