README.md

---
license: mit
language:
- nl
pretty_name: "ECLI → BWB legislation references (Rechtspraak / LIDO)"
size_categories:
- 10M<n<100M
configs:
- config_name: default
  data_files: ecli-bwb-id.parquet
---

# ECLI → BWB legislation references

Links Dutch case law (**ECLI**) to the Dutch legislation it cites (**BWB**
identifiers from the Basiswettenbestand register), resolved via LIDO
(Linked Data Overheid).

One row per (case, cited law element): a single ECLI has one row for every
statutory element it references.

## Stats

| | |
|---|---:|
| Rows | 10,198,164 |
| Distinct ECLIs | 766,079 |
| Distinct laws (bwb_id) | 7,112 |

Element types: `artikel` 8.3M · `wet` 1.8M · `afdeling` 35k · `hoofdstuk` 25k · `paragraaf` 6k · `deel` 820 · `boek` 431

## Columns

| Column | Type | Description |
|---|---|---|
| `legal_case_id` | int | Internal case id in the source database |
| `ecli` | string | European Case Law Identifier of the citing decision |
| `type` | string | Statutory element level: `wet`, `boek`, `deel`, `hoofdstuk`, `afdeling`, `paragraaf`, `artikel` |
| `bwb_id` | string | BWB identifier of the law (e.g. `BWBR0005290`) |
| `bwb_label_id` | int | BWB label id of the specific element |
| `jc_id` | string | JuriConnect deep reference (element + version date) |
| `number` | string | Element number within the law (e.g. article number) |
| `title` | string | Human-readable label (e.g. “Burgerlijk Wetboek Boek 7, Artikel 658”) |

## Usage

```python
import pandas as pd
df = pd.read_parquet("hf://datasets/davidwickerhf/ecli-bwb-id/ecli-bwb-id.parquet")

# All laws cited by one decision
df[df.ecli == "ECLI:NL:HR:2019:1734"]

# Most-cited articles
(df[df.type == "artikel"]
   .groupby(["bwb_id", "number", "title"]).size()
   .sort_values(ascending=False).head(20))
```

Deep links per row:
- `http://wetten.overheid.nl/{bwb_id}` — the law text
- The `jc_id` JuriConnect string pins the exact element and version date.