# sejm_api

Contemporary official Polish Sejm speeches, supplementing existing parliamentary sources.

## Dataset description

- Source (upstream): https://api.sejm.gov.pl/
- Source documentation: https://api.sejm.gov.pl/sejm.html
- Domain: political/spoken
- Language: Polish (`pl`)
- License: `public-domain (official documents)`
- Created (range): 2023-01-11, 2026-07-03
- Added: 2026-08-22
- Contributor: Piotr Styła (`PiotrSty` on Hugging Face; `PiotrStyla` on GitHub).

## Licensing — traceable basis

Official parliamentary documents and materials are excluded from copyright by
article 4(2) of the Polish Copyright Act: https://eli.gov.pl/api/acts/DU/2025/24/text/O/D20250024.pdf.

Reuse of public-sector information is governed by articles 2(12), 5, 6, 14,
15 and 17 of the Polish Open Data Act: https://eli.gov.pl/api/acts/DU/2023/1524/text.html.

The per-document value intentionally matches the existing `parliamentary`
source. No unverified upstream Creative Commons license is asserted.

## Provenance and reproducibility

- Official provider: Kancelaria Sejmu RP.
- Data route: Sejm API data via contributor snapshot. The text comes from the
  Hugging Face dataset `PiotrSty/sejm-speeches-corpus@v1.0.1`, not from the live
  API; SlayerLab did not re-fetch it from api.sejm.gov.pl.
- Immutable source release: `PiotrSty/sejm-speeches-corpus@v1.0.1`.
- Pinned source commit: `124557799608ce4212cf22fa90ca7904618de1b6`.
- Reviewed DynaWord base commit: `8cc48a3c564ac9edabfbbf012842ad11f70a3e5c`.
- Rebuild: `python src/fetch_sejm_api.py --help`, then `python3 src/clean_sejm_api.py`
  (trailer cut, below).
- Speaker attribution and source URLs are in `sejm_api.attribution.jsonl`. Each
  `source_url` is a day-level listing (`https://api.sejm.gov.pl/sejm/term{N}/videos/{date}`),
  not a per-speech URL.
- Slayer contribution digest: `sha256:38e42a791d68d593f82e1125f2bb94cd8f68212433e71d3e3da90f82d78eaf32`.

## Temporal novelty and deduplication

The Sejm material of `parliamentary` ends with term 8 (`2019-11-11`; its
Senate material is not compared: this source is Sejm-only). Every contributed
record is dated on or after `2023-01-01`, so its date range does not overlap it.
Exact duplicate speech texts within the contribution are removed using
SHA-1 fingerprints; this is not a claim of global cross-domain near-deduplication.

## Statistics

| Measure | Value |
|---|---:|
| Source rows inspected | 157,714 |
| Rows before 2023-01-01 | 118,902 |
| Statements split into rows of their own | 2,466 |
| Rows under 200 characters | 1,371 |
| Exact duplicate texts removed | 8 |
| Documents published | 39,899 |
| Characters | 83,482,628 |
| Tokens (`cl100k_base` proxy) | 30,860,439 |
| Documents with speaker attribution | 39,899 |

Statistics are recomputed from the released Parquet file. The canonical schema
is `id, text, source, added, created, token_count, license, author`.

## Trailer cut

The rights audit (slayerlabs/polish-dynaword-v2#57) found that the snapshot
appends to 34,980 of the 38,812 fetched speeches a "Przebieg posiedzenia" header
of a later statement of the same sitting day, and sometimes that statement's
text. `src/clean_sejm_api.py` cuts every trailer (23,485,670 characters). An
appended statement with text becomes a row of its own (2,466 rows), credited to
the speaker the header names ("Przebieg posiedzenia Janusz Kowalski: …"), without
the office in its label. Its date and term are those
of the row it was cut from: the snapshot labels the 11th sitting of term 10
(2024-05-08, 2024-05-09, 2024-05-15) "11. kadencja" in 696 headers. A split row's
id hashes its date, term, speaker and text like every other row; its sidecar
line carries `split_from` (the id of the row it was cut from in v0.2.5) and
`has_events: null`, since the snapshot's events
flag describes the whole parent row. Rows that fall under the fetch's
200-character minimum after the cut (1,318 cut speeches and 53 split rows) and 8
texts that became exact duplicates are dropped.

## Privacy and limitations

The upstream release heuristically redacts email addresses, PESEL identifiers
and unambiguous Polish telephone numbers. Public-official names and speech
attribution remain. Regex redaction is not a guarantee of zero personal data.
Downstream users must review applicable privacy, third-party and source terms.