README.md

metadata

license: cc-by-4.0
task_categories:
  - translation
language:
  - hi
  - as
  - bn
  - gom
  - gu
  - ks
  - mai
  - mr
  - mni
  - ne
  - or
  - pa
  - sat
  - sd
  - ta
  - te
  - ur
  - kn
  - ml
tags:
  - judiciary
  - legal
  - law
  - indian-languages
  - parallel-corpus
pretty_name: Judiciary_v2

Judiciary_v2: Hindi → Indian Languages Translation

Judiciary_v2 is a specialized parallel corpus for the Judiciary and Legal-Services domain. It encompasses court notices, case-status messages, legal aid information, and procedural guidelines across 18+ Indian languages.

📂 Folder Structure

Following the HIN-<LANG> convention, this dataset contains manually verified translations:

• Path: Judiciary_v2 / HIN-XXX / source_reviewed / JUD / *.txt
• Quality: All data is source-reviewed (manually translated and verified with QA).

📄 Data Format (TSV)

Tab-separated files with the following headerless structure: id | src_hi | tgt_<lang> | domain="judiciary"

Example: 1 न्यायालय ने फैसला सुरक्षित रख लिया है। <translated_sentence> judiciary

🌐 Supported Languages

Source: Hindi (hi) Targets: Assamese (as), Bangla (bn), Goan Konkani (gom), Gujarati (gu), Kashmiri (ks), Maithili (mai), Marathi (mr), Meitei (mni), Nepali (ne), Odia (or), Punjabi (pa), Santali (sat), Sindhi (sd), Tamil (ta), Telugu (te), Urdu (ur), Kannada (kn), Malayalam (ml).

⚖️ Data Sources

• Court and tribunal portals (public notices, orders, FAQs).
• Legal services authority materials and awareness brochures.
• e-Courts/e-filing templates and SMS/IVRS texts.

🚀 Intended Use

• NMT Training: Fine-tuning models for legal, judicial, and administrative terminology.
• Justice-Tech: Powering multilingual legal-aid chatbots and e-court information systems.