Research + corpus-building uses. Benchmarks don’t count.
The drawers
A tool can be filed in more than one drawer
DRAWER VIII
Public Datasets
Open corpora and collections others have built and shared. Each card says how the text was made and what you’re allowed to do with it.
Guide cards
Counts research and corpus-building uses in the window, plus “I’m using this now” check-ins. Benchmarks don’t count.
dataset · No. 102
Chronicling America
Library of Congress collection of millions of digitised US newspaper pages up to 1963, with freely downloadable OCR text. The default corpus for large-scale work on American newspapers.
At a glance
Holder
Library of Congress and NEH
Launched
2007
Coverage
US newspapers to 1963; about 24 million pages
Licence
Mostly public domain
Cost
Free
Access
Website; JSON API (no key); bulk OCR batches
Historian’s notes
How the text was madeUncorrected OCR, mostly from microfilm. One estimate puts word error above 25%.
Known gaps and biasesCoverage depends on which states took part and what each chose to digitise. Columns, headlines and ads are often scrambled together.
What you can do with itOlder issues are public domain; later ones may contain copyrighted material.
Nobody has said they’re using it in 2026 yet.
“I’m using this now” asks you to sign in with ORCID.
One count per person per year; it adds to Trending, and names are never shown.