license: mit
task_categories:
- text-generation
language:
- ko
tags:
- legal
pretty_name: Korean Statutes (All Laws of Korea)
size_categories:
- 100K<n<1M
π Korean Statutes Dataset (All Laws of Korea)
This dataset provides a comprehensive collection of 200,633 legal provisions from all Korean statutes, crawled from the Korea National Law Information Center (law.go.kr). Each entry corresponds to an individual clause (μ‘°λ¬Έ) within a law and includes metadata such as law name, promulgation date, enforcement date, issuing department, and more.
This dataset is ideal for legal NLP, statutory clause-level retrieval, and Korean law-aware generative AI systems.
π¦ Dataset Summary
- Total Records: 200,633
- Total Laws: 5,500
- Language: Korean (
ko) - Fields: 15 columns with English names
- Source: law.go.kr
- Contact: ducut91@gmail.com
- 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.
Each row corresponds to one clause (
μ‘°λ¬Έ) from a law. The full law content can be retrieved via the providedLawDetailLink.
π§Ύ Column Descriptions
| Column Name | Description |
|---|---|
CurrentHistoryCode |
Status of the law version (e.g., "νν") |
LawSerialNumber |
Unique identifier for the law version |
LawDetailLink |
Direct API link to the full law text (in XML format) |
LawNameKor |
Full name of the law in Korean |
LawTypeName |
Type of law (e.g., "λ²λ₯ ", "λν΅λ Ήλ Ή") |
DepartmentName |
Name of the issuing or managing ministry/agency |
PromulgationNumber |
Official number assigned at promulgation |
RevisionTypeName |
Revision type (e.g., "μΌλΆκ°μ ", "μ λΆκ°μ ", "νλ²κ°μ ") |
LawID |
Unique identifier for the law itself (not version-specific) |
EffectiveDate |
Date the law goes into effect (YYYY-MM-DD) |
PromulgationDate |
Date the law was promulgated (YYYY-MM-DD) |
LawAbbreviation |
Short or abbreviated name of the law |
ClauseNumber |
Clause number (e.g., "2", "3μ2", "5μ‘°μ3"), stored as a string |
ClauseTitle |
Title of the clause (if present) |
ClauseContent |
Full text content of the clause, including subparagraphs and line breaks |
π Missing Value Statistics
| Column Name | Missing Count | Missing Ratio (%) |
|---|---|---|
DepartmentName |
130 | 0.06% |
LawAbbreviation |
117,379 | 58.50% |
ClauseTitle |
12,333 | 6.15% |
| Other columns | 0 | 0.00% |
β οΈ The
LawAbbreviationfield is missing in over half the entries, but this does not affect core clause data.
π οΈ Preprocessing Notes
- Crawled using the DRF Open API provided by law.go.kr.
- Each
LawDetailLinkis a direct link to the XML representation of the full law version. - Long multiline clauses (including subparagraphs and notes) are preserved in
ClauseContent.
π‘ Suggested Use Cases
- π§ Statute Summarization β Train models to summarize Korean legal texts
- π€ Legal QA and Chatbots β Power clause-aware legal assistant systems
- π Legal Language Modeling β Use for pretraining or fine-tuning LLMs
- π Structured Legal Retrieval β Build clause-level retrieval or embedding systems
π License & Purpose
This dataset is based entirely on public legal data from the Republic of Korea. According to Korean copyright law, laws and other legal texts are not subject to copyright, and may be freely:
- Used
- Modified
- Redistributed
for research, commercial, or educational purposes.
While no attribution is required, citing law.go.kr as the source is recommended.
π― Motivation for Open-Sourcing
We believe that Korean law should be as accessible to developers and researchers as the laws of other major jurisdictions. However, many large language models underrepresent Korean legal content.
By releasing this dataset, we aim to:
- Promote equitable legal AI development
- Encourage multilingual legal research
- Enable the next generation of Korean legal tech
π°π· A globally aware model needs to understand Korean law.
π Changelog
| Date | Changes Made |
|---|---|
| 2025-07-25 | Uploaded first version. |
| 2025-07-27 | Removed structural items (νΈΒ·μ₯Β·μ Β·κ΄) from ClauseContent; updated statistics. |