Where the corpus lives.
MEDIATEXT is distributed in two layers: a public reading portal for the V4 release, and a Harvard Dataverse + OSF deposit for the full programmatic data. This page lists the published sub-corpora and the country-level browsing surfaces sourced from the same coded data.
Flagship sub-corpora
The two releases documented in companion data papers (currently in editorial review), plus the historical US-media surface that predates the current codebook.
MediaText CEE (V4)
1.4M+ articles · 1990–2024 · sentence-level · CC BY-NC 4.0
MediaText NYT
1851–present · closed period 1851–2024 · two access tiers
USA media (historic)
NYT / WaPo / etc. · earlier coding vintage · pre-MEDIATEXT
MediaText EU
Brussels press + member-state spillover · planned release
Country browsing surfaces
Companion country-level browsing portals sourced from the same coded data — same codebook, same label schema, same sentence-level granularity. These complement the main MEDIATEXT release: handy for inspecting individual outlets without downloading the full deposit. Click a card to open the corresponding country surface.