README.md

metadata

pretty_name: ECHR Guide Case Catalog
language:
  - en
tags:
  - legal
  - echr
  - hudoc
  - entity-linking
  - dataset-creation
size_categories:
  - 1K<n<10K

Dataset Card for ECHR Guide Case Catalog

Dataset Summary

This dataset links citation strings found in ECHR guide materials to canonical HUDOC case records and guide-level appearances.

It is derived from the cleaned citation-diff layer and a HUDOC enrichment pass. The goal is to move from raw citation strings to normalized case entities with stable identifiers, document types, importance levels, and guide appearance metadata.

Dataset Structure

Main files:

• cases_catalog.csv: main case catalog table
• cases_catalog.json: JSON export of the same catalog
• case_guides.csv: guide-level case associations
• case_appearances.csv: finer-grained appearance-level mapping
• audit/cases_catalog_hudoc_report.json: matching statistics and audit counts
• audit/cases_catalog_hudoc_unmatched.csv: unresolved citation rows

Dataset Statistics

• catalog rows: 7846
• matched rows: 7763
• unmatched rows: 83
• guide-link rows in case_guides.csv: 12310
• appearance rows in case_appearances.csv: 61646

Intended Use

This dataset is intended for:

• canonical case linking
• joining guide citations to HUDOC metadata
• filtering or ranking guide changes by case importance or document type
• serving as an intermediate layer between citation diffs and body-context localization

Provenance

Derived from:

• lexgenie/echr-guide-citation-diffs
• the cleaned citation-diff outputs in lexgenie/echr-guide-citation-diffs-cleaned
• HUDOC enrichment and matching

This dataset should be treated as a normalized entity layer rather than as a raw citation-diff release.