Datasets:
Add CC BY-SA 3.0 Polish media-literacy lessons
Add Polish media-literacy lessons
Adds edukacja_medialna_pl: 252 records and 728,987 measured cl100k_base proxy tokens, acquired directly from official Edukacja Medialna lesson XML. 255 lesson URLs discovered; 3 dropped as missing_creator. Every retained record has explicit CC BY-SA 3.0 evidence on its lesson page and in the source XML, plus credited authors (100% coverage, 62 distinct authors).
Source dataset: https://huggingface.co/datasets/PiotrSty/edukacja-medialna-pl/tree/ba738c60fe1a92a676a6cabba05e8c3f82e125b7
Provenance and license
- Origin: Edukacja Medialna (Fundacja Nowoczesna Polska), lesson index https://edukacjamedialna.edu.pl/lekcje/. Per-lesson source XML (
utworwith Dublin Core fields and the authoredpowiescbody) is fetched and checksummed (page_sha256,xml_sha256), pinned inartifacts/raw_xml.jsonl.gz. - License statement at the origin (the mirror is not evidence for itself):
- every lesson page ends with
Licencja: Creative Commons Uznanie autorstwa - Na tych samych warunkach 3.0, linking https://creativecommons.org/licenses/by-sa/3.0/ — e.g. https://edukacjamedialna.edu.pl/lekcje/abc-bezpieczenstwa-w-sieci/; - the site footer repeats the grant: "Jeśli nie oznaczono inaczej, wszystkie materiały na stronie są objęte wolną licencją Creative Commons Uznanie autorstwa – Na tych samych warunkach 3.0" (http://creativecommons.org/licenses/by-sa/3.0/deed.pl);
- the source XML carries
dc:rights+dc:rights.license.
- every lesson page ends with
- Build gate: explicit CC BY-SA 3.0 required on page AND XML. Per-record
rights/rights_urlsand page/XML/text hashes inartifacts/attribution.jsonl.
What the texts are
- Genre: authored teaching materials — knowledge explainers ("Wiedza w pigułce"), lesson plans and projects ("Pomysł na lekcję", "Zadanie": objectives, classroom procedure, evaluation), glossary entries ("Słowniczek"). One document per lesson.
- Register: prepared didactic Polish — formal written prose with classroom instructions, exercises and check questions; gender-inclusive paired forms ("uczestnicy i uczestniczki"). Human-authored, not OCR'd, not machine-translated; 62 credited authors (roles
textbook,scenario,expert,methodologist). - Time span: upstream publication dates
2012-11-09–2015-01-12.
Collection and normalization (all in src/build_edukacja_medialna_pl.py)
- Extract only the authored
powiescbody (leaf blocks); extraction stops at the "Czytelnia" heading — external reading lists, bibliography and long-quote elements are omitted. - Normalize: NFKC; soft hyphens (U+00AD), zero-width spaces (U+200B) and control characters removed; space/tab/NBSP runs collapsed per line, lines trimmed; runs of 3+ newlines capped at one empty line (paragraph breaks kept).
- PII: pattern-based redaction only — e-mail →
[REDACTED:EMAIL], labelled phones (telefon/tel.) →[REDACTED:PHONE]. Matches in the shard: 1 e-mail, 0 phones. This is not de-identification; a NERGAL pass is required before release. - Gates and drops: license explicit on page+XML (255/255), credited creator (−3,
missing_creator), body ≥ 300 chars (−0), langid Polish (7-language set, confidence ≥ 0.99; −0), exact dedup on normalized text (−0), near-dedup seeded MinHash 128 / LSH 0.8 with exact 5-word Jaccard ≥ 0.9 (−0). 252 of 255 kept.
Before/after — extraction and normalization ate no real content:
Raw XML → built text (entity decoded, inline markup flattened, link target dropped):
RAW XML
<punkt>rozumieją, że znajomość możliwie wielu konwencji gatunkowych jest uważana za cechę świadczącą o wysokiej kulturze osobistej? (--> Rozważ realizację lekcji <link url="http://edukacjamedialna.edu.pl/lekcje/medialny-savoir-vivre">,,Medialny savoir-vivre"</link>.)</punkt>
TEXT
rozumieją, że znajomość możliwie wielu konwencji gatunkowych jest uważana za cechę świadczącą o wysokiej kulturze osobistej? (--> Rozważ realizację lekcji ,,Medialny savoir-vivre" .)
Reading-list cut (raw XML continues below; the built text of edukacja_medialna_pl_abc-bezpieczenstwa-w-sieci stops after "Słowniczek"):
RAW XML
<naglowek_rozdzial>Czytelnia</naglowek_rozdzial>
<lista typ="czytelnia">
<punkt><tytul_dziela>Jak dzieci i młodzież chronią swoją prywatność?</tytul_dziela>, GIODO [dostęp: 23.06.2013]: <link url="http://www.giodo.gov.pl/560/id_art/4699/j/pl/">http://www.giodo.gov.pl/560/id_art/4699/j/pl/</link></punkt>
TEXT (tail)
Słowniczek
anonimowość
media społecznościowe
nick
prywatność
geolokalizacja
CC BY-SA 3.0 at corpus level — proposal
- Attribution, per record:
authorin the parquet plusartifacts/attribution.jsonl—authorswithcreator_roles,title,canonical_url/page_url/xml_url,publisher,created,license+license_url, sourcerights/rights_urls,page_sha256/xml_sha256/text_sha256. Enough to credit each document under BY-SA 3.0 §4 without consulting the mirror. - Share-alike note (now in the datasheet): the corpus is an adaptation and is released under CC BY-SA 4.0 — a later version of the same License Elements, which BY-SA 3.0 §4(b) permits for adaptations. Redistributors of these records or adaptations of them must keep the per-record attribution and pass share-alike terms on (CC BY-SA 3.0 or later, same License Elements).
Open to the maintainer's preferred wording for the corpus-level treatment.
Sample (verbatim, python3 src/render_sample.py data/edukacja_medialna_pl/edukacja_medialna_pl.parquet -n 3)
id: edukacja_medialna_pl_standardy-dostepnosci
source: edukacja_medialna_pl
added: 2026-09-06
created: 2012-11-09
token_count: 2150
license: CC-BY-SA-3.0
author: Baranowska, Joana Ruta; Paszewska, Weronika; Zadrożny, Jacek
Standardy dostępności
Wiedza w pigułce
W 2012 roku Polska ratyfikowała Konwencję ONZ o prawach osób niepełnosprawnych. Przywołuje ona Powszechną Deklarację Praw Człowieka i przypomina o tym, że wszyscy mają równe prawa oparte o poszanowanie godności każdej jednostki. Ze względu na ograniczenia osób niepełnosprawnych te prawa muszą być po prostu inaczej realizowane.
Wytyczne dotyczące dostępności treści internetowych
Web Content Accessibility Guidelines
12 wytycznych podzielono na 4 grupy odpowiadające czterem zasadom dostępności treści:
zrozumiałość --- teksty są czytelne i łatwe w odbiorze; interfejs jest intuicyjny i logicznie zbudowany; trudno popełnić błąd przy jego obsłudze
bez przeszkód w obsłudze --- strona jest możliwa w obsłudze przy użyciu klawiatury; nie ma limitów […]
id: edukacja_medialna_pl_wizerunek-w-sieci
source: edukacja_medialna_pl
added: 2026-09-06
created: 2012-11-09
token_count: 4613
license: CC-BY-SA-3.0
author: Dobrowolska, Urszula; Muszyński, Damian; Wysocki, Michał
Wizerunek w sieci
Wiedza w pigułce
Zwykle przejmujemy się tym, co myślą o nas inni. Opinia na nasz temat zależy od wielu czynników. Może mieć na nią wpływ nasze zachowanie, ubiór czy zainteresowania. Wszystko to składa się na nasz wizerunek.
We współczesnym świecie internet stał się ważną przestrzenią tworzenia naszego wizerunku. Nasze aktywności w sieci mają na niego duży wpływ.
Na internautach leży odpowiedzialność za to, co udostępnią. Nawet po usunięciu pewnych treści, w sieci pozostaje po nich ślad, np. na serwerach wyszukiwarek albo jeśli ktoś wcześniej je skopiuje. Dlatego też powinieneś zwracać szczególną uwagę na to, co i jak mówisz w internecie. Na twój wizerunek w sieci wpływają m. in.:
twój sposób wyrażania się --- to, czy piszesz zgodnie z zasadami poprawnej polszczyzny, […]
id: edukacja_medialna_pl_matura-i-co-dalej
source: edukacja_medialna_pl
added: 2026-09-06
created: 2013-04-17
token_count: 5459
license: CC-BY-SA-3.0
author: Jasiewicz, Justyna; Wysocki, Michał
Matura i co dalej?
Zadanie
Projekt zakłada przeprowadzanie przez poszczególne grupy projektowe analizy szans, jakie na rynku pracy daje ukończenie różnych kierunków studiów.
Celem projektu jest:
rozwój umiejętności wyszukiwania informacji, a także przeprowadzania ich analizy i syntezy;
wzrost kompetencji w zakresie umiejętności pracy w zespole;
zachęcenie do podjęcia świadomej decyzji dotyczącej dalszej edukacji.
Zakres podjętej analizy może zostać dostosowany do zainteresowań grupy. Uczestnicy i uczestniczki mogą badać między innymi: szanse znalezienia pracy po kierunkach ścisłych lub humanistycznych, losy absolwentów konkretnej uczelni lub kierunku studiów lub zdobyć informacje na temat kierunku studiów, który gwarantuje pracę w danej branży.
Rezultat
Poszczególne grupy […]
Inclusion argument and caveats
What this adds: authored didactic prose — objectives, exercises, classroom procedure, glossary definitions — a school-teaching register the corpus does not otherwise draw from (encyclopedic, legal, parliamentary, journalistic, fiction). 728,987 tokens is a diversity addition, not a scale one.
Caveats: single publisher and 62 credited authors; school-audience framing with gender-inclusive paired forms; technology/legal guidance frozen at 2012–2015; embedded third-party excerpts (quotations, linked works) may need separate review. Training benefit is an untested diversity hypothesis.
Remaining gates
NERGAL pass before release (in-build redaction is pattern-only). Cross-source exact/near deduplication and benchmark-overlap checks remain pending. Historical technology/legal guidance is not asserted to be current. This proposes a source (release: None), not a stable release.
Test plan
-
python3 -m pytest src/test_edukacja_medialna_contribution.py(2 tests) - Reviewer: texts are didactic lesson prose (explainer + scenario), reading lists absent
- Reviewer: origin license statement accepted as evidence
- Reviewer: corpus-level BY-SA proposal accepted or amended
Review against AGENTS.md ("What the PR must show", "Normalize the text before you build"). Needed before merge:
- Rebase onto current
main. Against main the diff currently showseuropean_hplt_v3_plbeing re-added (removed from main for unverified license) anduzp_orzeczenia_pl/rock_pollub_plbeing deleted.git fetchof this PR's ref currently fails on our side; a fresh push after the rebase should fix it. - Drop
src/scrub_pii.py/src/scrub_entities.py. In-repo scrubbers were removed in #38; PII goes through SlayerLab/NERGAL. Merge withrelease: Noneis fine meanwhile, but no release before a NERGAL pass. - Keep the PR to this source. Drop edits to README / CHANGELOG /
VERSIONin make_docs and to shared files (AGENTS.md,clean_hplt_v3.py, the sejm fetchers). Registry entry stays, withrelease: None; the release bookkeeping is done once by the maintainer after merges. - License at the origin. Link the license statement on the Edukacja Medialna site (lesson page), not only your HF mirror — the mirror can't be evidence for itself.
- BY-SA. You leave the corpus-level treatment of CC BY-SA 3.0 open — propose it (attribution fields per record, share-alike note in the datasheet).
- Describe the texts — genre, register, time span.
- Sample inline. Paste 2–3 real documents into the description, one fenced block each: metadata as
key: valuelines, a blank line, then the text with its real line breaks, truncated with[…](format as in #24). A linked.sample.jsonlrenders on the Hub as one escaped line per document and can't be reviewed at a glance.
Correction to my review comment above: the git fetch failure was on the Hub side (have/ack negotiation), not a problem with this PR. It fetches fine with git -c fetch.negotiationAlgorithm=noop fetch. No re-push needed for that; the rebase ask still stands.
Review items addressed — pushed as three commits on refs/pr/26 (d6e04ec, aba24aa, d85848e, rebased onto main @ 34e7ea3):
- Rebased onto current
main. The two-dot diff is now scoped to this source: noeuropean_hplt_v3_plre-add, nouzp_orzeczenia_pl/rock_pollub_pldeletions. Thesrc/sources.pyconflict resolved to one new entry (20 added lines, nothing else). The earliergit fetchfailure is the Hub-side have/ack issue you corrected — usedgit -c fetch.negotiationAlgorithm=noopper AGENTS.md. - Scrubbers gone.
src/scrub_pii.py/src/scrub_entities.pywere never in this PR's commits (branch predates #38); the rebase removes them from the diff. PII story now documented in the datasheet: pattern redaction in the builder (1 e-mail, 0 phones) and no release before a NERGAL pass. Registry entry kept withrelease: None. - PR scoped to this source. Only
data/edukacja_medialna_pl/*,src/build_edukacja_medialna_pl.py,src/test_edukacja_medialna_contribution.py, the registry entry, and an appendedartifacts/source_findings.mdentry. README / CHANGELOG /VERSION/RELEASE_DATEin make_docs, AGENTS.md,clean_hplt_v3.pyand the sejm fetchers untouched. Side note:attribution.jsonl/decisions.jsonlwere raw blobs from the pre-#38.gitattributes; they are now proper LFS pointers per the current rule. - License at the origin. NOTICE.md, datasheet and the description now cite the origin's own statement: the per-lesson
Licencja: Creative Commons Uznanie autorstwa - Na tych samych warunkach 3.0line (e.g. https://edukacjamedialna.edu.pl/lekcje/abc-bezpieczenstwa-w-sieci/) and the site footer (http://creativecommons.org/licenses/by-sa/3.0/deed.pl), plus the XMLdc:rights/dc:rights.licensefields. The mirror is referenced only as the pinned snapshot. - BY-SA proposal written into the datasheet ("Corpus-level CC BY-SA 3.0 treatment"): per-record attribution fields (
author+artifacts/attribution.jsonlwith roles, canonical URLs, rights evidence and hashes) and a share-alike note — the corpus is an adaptation released under CC BY-SA 4.0, a later version of the same License Elements per BY-SA 3.0 §4(b), with attribution and share-alike passed on. Happy to use the maintainer's preferred wording. - Texts described (description + datasheet): genre (lesson explainers, lesson plans/projects, glossary entries), register (prepared didactic Polish, gender-inclusive paired forms, human-authored), time span (2012-11-09 – 2015-01-12).
- Sample inline in the description — 3 real documents (
render_sample.py -n 3, one fenced block each:key: valuemetadata, blank line, text with real line breaks,[…]truncation), plus two before/after excerpts of the extraction/normalization.
Also reported per AGENTS.md "Normalize the text before you build": every normalization step lives in src/build_edukacja_medialna_pl.py, and the datasheet now lists the steps, the gates and what they dropped (3 × missing_creator; all other gates −0).
Verification: python3 -m pytest src/test_edukacja_medialna_contribution.py — 2 passed.
refs/pr/26 ref