Structured Dataset of Egyptian Drug-Related Criminal Cases
Description
This dataset contains structured legal data extracted from multiple Egyptian legal references, primarily focused on drug-related criminal cases. The sources include the books "مجموعة الأحكام الصادرة من الهيئة العامة للمواد الجنائية ومن الدوائر الجنائية" (Collection of Judgments Issued by the General Authority for Criminal Matters and Criminal Circuits) and "الموسوعة الذهبية في قضايا المخدرات" (The Golden Encyclopedia of Drug Cases), both of which are publicly available and government-authorized. The dataset includes approximately 200 entries and 9 columns, covering key fields such as charges, case facts, legal reasoning, applicable laws, judgment outcomes, prison terms, fines, and more. Data was extracted using a combination of manual transcription and OCR techniques, and then formatted into a structured CSV file. This resource is designed for researchers and NLP engineers working in Arabic legal text processing. It supports various tasks including classification, prediction (e.g., law prediction based on facts), legal reasoning analysis, and fine-tuning Arabic legal NLP models. Use Restrictions: This dataset is intended strictly for academic and research purposes. It must not be redistributed for commercial use or re-published under another name. Any use must acknowledge the original source and contributors.
Files
Steps to reproduce
Review Source Materials Consult the official Egyptian legal books used to create the dataset: "مجموعة الأحكام الصادرة من الهيئة العامة للمواد الجنائية ومن الدوائر الجنائية" "الموسوعة الذهبية في قضايا المخدرات" Case Selection Identify drug-related criminal rulings relevant to the dataset focus. Text Extraction Extract case content using manual transcription or OCR (Optical Character Recognition). Data Structuring Organize each case into the following fields: Id, Charge, Facts, GroundReasons, Laws, Judgment, Prison, Fine, Class/Description. Data Cleaning Fix OCR errors, ensure consistent formatting, and remove personal identifiers. Save as CSV Store the structured and cleaned data in CSV format with UTF-8 encoding for easy reuse in NLP tasks.