constcourtPL / README.md
lszoszk's picture
Add case_metric (Podmiot/Przedmiot/Wzorce) to all 3672 cases
61cfe12 verified
---
language:
- pl
license: other
pretty_name: Trybunal Konstytucyjny Full Corpus
tags:
- legal
- constitutional-law
- polish
- retrieval
---
# Trybunal Konstytucyjny Full Corpus
This dataset repository stores a full snapshot of normalized decisions used by the `trybunalkonstytucyjny` dashboard.
## Snapshot metadata
- HF dataset repo: `lszoszk/constcourtPL`
- Source repository: `https://github.com/lszoszk/trybunalkonstytucyjny`
- Active profile: `full`
- Total files uploaded: 32
- Total payload size: 904.32 MB
- Cases: 3672
- Paragraphs: 284991
- Date range: 1986-2026
- Dataset hash: `413c82b97a083afc98198d5997f4296529d22d21c9aab572fe6abadab8b1963f`
- Dataset generated at: `2026-02-25T14:07:05.345Z`
- Normalization version: `tk-norm-v2`
- Includes sample datasets: no
## What this dataset contains
- `data/tk_cases.jsonl`: full normalized corpus of Polish Constitutional Tribunal decisions (metadata + paragraph-level text units).
- `data/tk_cases_full_bench.jsonl`: subset of decisions resolved in full bench.
- `data/similar_cases*.json`: precomputed nearest-case links for retrieval and exploration.
- `data/stats.json`: aggregate counts, rankings, date coverage, and dataset fingerprint metadata.
- `data/lemma_shards/*`: lemma/form/POS shards used for lemma-aware search in the expert dashboard.
## Application
- Live app (landing): https://lszoszk.github.io/trybunalkonstytucyjny/
- Expert search dashboard: https://lszoszk.github.io/trybunalkonstytucyjny/docs/dashboard-expert.html
- Analytics dashboard: https://lszoszk.github.io/trybunalkonstytucyjny/docs/analytics.html
## Files
- `data/lemma_shards/full/forms-001.json` (459.64 KB)
- `data/lemma_shards/full/forms-002.json` (554.51 KB)
- `data/lemma_shards/full/forms-003.json` (596.31 KB)
- `data/lemma_shards/full/forms-004.json` (578.21 KB)
- `data/lemma_shards/full/forms-005.json` (594.13 KB)
- `data/lemma_shards/full/forms-006.json` (571.49 KB)
- `data/lemma_shards/full/forms-007.json` (325.32 KB)
- `data/lemma_shards/full/lemma-pos-001.json` (52.54 MB)
- `data/lemma_shards/full/lemma-pos-002.json` (74.95 MB)
- `data/lemma_shards/full/lemma-pos-003.json` (88.65 MB)
- `data/lemma_shards/full/lemma-pos-004.json` (66.61 MB)
- `data/lemma_shards/full/lemmas-001.json` (23.55 MB)
- `data/lemma_shards/full/lemmas-002.json` (32.92 MB)
- `data/lemma_shards/full/lemmas-003.json` (39.36 MB)
- `data/lemma_shards/full/lemmas-004.json` (28.59 MB)
- `data/lemma_shards/full/manifest.json` (1.77 KB)
- `data/lemma_shards/full_bench/forms-001.json` (504.73 KB)
- `data/lemma_shards/full_bench/forms-002.json` (571.76 KB)
- `data/lemma_shards/full_bench/forms-003.json` (582.68 KB)
- `data/lemma_shards/full_bench/forms-004.json` (519.89 KB)
- `data/lemma_shards/full_bench/lemma-pos-001.json` (24.96 MB)
- `data/lemma_shards/full_bench/lemma-pos-002.json` (30.79 MB)
- `data/lemma_shards/full_bench/lemma-pos-003.json` (5.85 MB)
- `data/lemma_shards/full_bench/lemmas-001.json` (10.43 MB)
- `data/lemma_shards/full_bench/lemmas-002.json` (12.68 MB)
- `data/lemma_shards/full_bench/lemmas-003.json` (2.38 MB)
- `data/lemma_shards/full_bench/manifest.json` (1.27 KB)
- `data/similar_cases.json` (11.56 MB)
- `data/similar_cases_full_bench.json` (1.27 MB)
- `data/stats.json` (24.43 KB)
- `data/tk_cases.jsonl` (311.49 MB)
- `data/tk_cases_full_bench.jsonl` (80.00 MB)
## Notes
- Main training/evaluation source: `data/tk_cases.jsonl`
- Full-bench subset: `data/tk_cases_full_bench.jsonl`
- Lemma shards are under `data/lemma_shards/full*`