metadata

license: mit
task_categories:
  - text-generation
language:
  - ko
tags:
  - legal
pretty_name: korea_statute_interpretations
size_categories:
  - 1K<n<10K

📚 Korean Statute Interpretations Dataset

This dataset contains 8,344 official statute interpretations published by the Korea Ministry of Government Legislation (법제처). Each record represents an authoritative interpretation of Korean statutes in response to administrative inquiries, making this dataset valuable for legal NLP, regulatory reasoning, and generative legal AI applications.

📦 Dataset Summary

Total Records: 8,344
Language: Korean (ko)
Fields: 14 columns
Source: law.go.kr
Contact: [email protected]
License: This dataset is based on public legal data provided by the National Law Information Center (law.go.kr) and is freely available for anyone to use. You may use, modify, and redistribute the data for research, commercial, or educational purposes without restriction.

All data was collected from the official Korea Law Information Service using their public API and document pages.

🧾 Column Descriptions

Column Name	Description
`Title`	Title of the interpretation inquiry
`DetailLink`	URL to the full interpretation document in HTML format
`ReplyOrgName`	Organization that issued the reply (usually 법제처)
`CaseNumber`	Identifier used by the agency for internal management
`CaseSerialNumber`	Unique serial number of the case
`ReplyOrgCode`	Numerical code of the responding organization
`InquiryOrgName`	Name of the agency that submitted the inquiry
`ReplyDate`	Date when the reply was issued (format: YYYY.MM.DD)
`InquiryOrgCode`	Code representing the inquiring institution
`Summary`	Legal background and explanation of the case context
`Answer`	Official response to the legal question
`Reason`	Legal and regulatory reasoning supporting the answer
`RelatedCases`	Related judicial precedents cited in the reasoning
`Etc`	Legal disclaimer about the scope and binding force of the interpretation (non-binding government advice)

📊 Missing Value Statistics

Column Name	Missing Count	Missing Ratio (%)
`InquiryOrgName`	1,002	12.01%
`ReplyDate`	91	1.09%
`InquiryOrgCode`	3,476	41.66%
`RelatedCases`	3,497	41.91%
Other Columns	0	0.00%

⚠️ RelatedCases and InquiryOrgCode have high missing rates, but core columns like Summary, Answer, and Reason are complete.

💡 Suggested Use Cases

💬 Legal Q&A Systems – Answer citizen questions by retrieving similar prior interpretations
🧠 Legal Reasoning Models – Train models to understand statutory interpretation patterns
🛠️ RAG-based Legal Assistants – Use in retrieval-augmented generation pipelines
📊 Policy & Regulation Mining – Analyze trends in public law application across agencies

🛠️ Preprocessing Notes

All texts are kept in Korean, preserving legal nuances.
The dataset was not modified or AI-augmented – it represents the official government-published content verbatim.
Long text fields (Summary, Answer, Reason) are especially useful for long-form legal understanding tasks.

📜 License & Use

This dataset is derived from Korean government open data and is not subject to copyright under Korea’s law.

You are free to:

Use, copy, modify, and distribute for research, educational, or commercial purposes
Incorporate into LLM training, retrieval systems, or benchmarking datasets

Attribution to law.go.kr is appreciated but not legally required.

🎯 Why We Open-Sourced This

We believe global legal language models should incorporate the Korean legal system, including statutory interpretation, which plays a critical role in public administration.

By releasing this dataset, we aim to:

Support multilingual legal AI research
Promote transparent legal reasoning models
Provide high-quality, structured Korean legal content to the open-source community

🇰🇷 Understanding the law isn't just for lawyers. Let’s build systems that help everyone.

📅 Changelog

Date	Changes Made
2025-07-24	First release with 8,344 interpretations and 14 structured columns.