sukhrobnurali/uzbek_constitution

Constitution of the Republic of Uzbekistan (Trilingual Dataset)

Dataset Summary

This dataset contains the Constitution of the Republic of Uzbekistan (New Edition, adopted via referendum on April 30, 2023) aligned in three languages:

• 🇺🇿 Uzbek (Latin script)
• 🇷🇺 Russian
• 🇬🇧 English

The data was carefully scraped and processed from the official National Database of Legislation of the Republic of Uzbekistan. It serves as a high-quality parallel corpus for legal NLP tasks, machine translation, and constitutional analysis.

Dataset Versions

This dataset is available in two structural variants to support different use cases.

1. Default Version (Original Schema)

• File: data/constitution.jsonl
• Description: In this version, Chapter and Part titles are single strings that include both the numbering (Roman/Word) and the title text.
• Fields: article_number, text_{lang}, chapter_{lang}, part_{lang}.

Sample Data (V1):

{
  "article_number": "1",
  "text_uz": "Oʻzbekiston — boshqaruvning respublika shakliga ega boʻlgan suveren, demokratik, huquqiy, ijtimoiy va dunyoviy davlat...",
  "chapter_uz": "I bob. Davlat suvereniteti",
  "part_uz": "BIRINCHI BOʻLIM. ASOSIY PRINSIPLAR",
  "text_ru": "Узбекистан — суверенное, демократическое, правовое, социальное и светское государство...",
  "chapter_ru": "Глава I. Государственный суверенитет",
  "part_ru": "РАЗДЕЛ ПЕРВЫЙ. ОСНОВНЫЕ ПРИНЦИПЫ"
}

2. Alternative Version (Structured Schema)

• File: data/constitution_v2_different_table_structure.jsonl
• Description: This version provides granular metadata. The Chapter and Part numbering are separated into their own integer columns, leaving the text columns clean.
• Fields: article_number, text_{lang}, chapter_number, chapter_{lang} (text only), part_number, part_{lang} (text only).

Sample Data (V2):

{
  "article_number": 1,
  "text_uz": "Oʻzbekiston — boshqaruvning respublika shakliga ega boʻlgan suveren, demokratik, huquqiy, ijtimoiy va dunyoviy davlat...",
  "part_number": 1,
  "part_uz": "ASOSIY PRINSIPLAR",
  "chapter_number": 1,
  "chapter_uz": "Davlat suvereniteti",
  "part_en": "FUNDAMENTAL PRINCIPLES",
  "chapter_en": "State Sovereignty"
}

Usage

You can load specific versions of the dataset using the data_files argument.

Loading with Hugging Face Datasets

from datasets import load_dataset

# Option 1: Load the Default Version (V1)
dataset_v1 = load_dataset("sukhrobnurali/uzbek_constitution", data_files="data/constitution.jsonl")
print(dataset_v1["train"][0]["chapter_uz"]) 
# Output: "I bob. Davlat suvereniteti"

# Option 2: Load the Structured Version (V2)
dataset_v2 = load_dataset("sukhrobnurali/uzbek_constitution", data_files="data/constitution_v2_different_table_structure.jsonl")
print(dataset_v2["train"][0]["chapter_number"]) 
# Output: 1

Applications

• Machine Translation: Parallel legal text for training/finetuning translation models (Uzbek-English, Uzbek-Russian).
• RAG (Retrieval-Augmented Generation): Reliable ground truth for legal Q&A bots.
• Legal Analysis: Comparative studies of constitutional terminology across languages.

License

This dataset is a derivative work of public legal documents. According to the Law of the Republic of Uzbekistan "On Copyright and Related Rights" (Article 8), official documents (laws, resolutions, decisions, etc.) and their official translations are not objects of copyright.