description string | samples list |
|---|---|
this is Samples of 250 deep meaning, widly used arabic toxic, sexual harasment, hate words written in arabizi : | [
{
"arabizi": "Enta mo6a5allef",
"english_translation": "you mental retardation",
"category": "harassment",
"label": "toxic"
},
{
"arabizi": "Ente Ka7beh",
"english_translation": "you are a whore",
"category": "sexual_harassment",
"label": "toxic"
},
{
"arabizi": "Koss omm... |
license: other task_categories: - text-classification task_ids: - hate-speech-detection language: - ar tags: - arabizi - dialectal-arabic - toxicity - harassment - social-media-safety - red-teaming size_categories: - n<1K
🛡️ Pro-Grade Arabizi 250- Sentences - Safety & Moderation Dataset
Purchase Full Access
👉 Buy the Complete Dataset on Gumroad Arabizi Toxicity & Content Moderation Dataset A dataset containing Arabizi (3araby) toxic, abusive, and profane terms used for training AI content moderation systems, Large Language Model (LLM) safety guardrails, and dialectal Arabic NLP pipelines.
*Access to the full dataset.json with Link Gumrod.
Dataset Summary
This is a premium, highly specialized dataset containing Modern Standard Arabic, regional dialects, and 250 Arabizi (Arabic written in Latin script with numbers) expressions curated specifically for content moderation. It focuses on the most critical safety categories standard in modern AI moderation platforms. ONERAHIP & COPYRIGHT DISCLAIMER
🛡️ Data Ownership, Origin & Commercial License Guarantee
- Original Creation & Expert Curation: This dataset was natively created, researched, and curated entirely by an expert native Arabic linguist specializing in regional dialects, Arabizi (3araby), and content safety.
- Full Legal Ownership: The creator guarantees 100% legal ownership and intellectual property rights over this dataset. It contains no proprietary third-party text, scraped copyright materials, or unauthorized private data.
- Commercial Usage Rights: Purchasing or acquiring this dataset grants developers, AI researchers, and commercial entities a non-exclusive, perpetual, worldwide license to integrate, fine-tune, and utilize this data in commercial and non-commercial Artificial Intelligence, Large Language Models (LLMs), and Safety Guardrail systems.
🚀 Custom Arabic & Arabizi AI Consulting & Datasets
Need custom data, evaluation, or safety testing for your Arabic AI models? As a native Arabic & Arabizi language expert, I offer the following professional services:
- AI Red Teaming & Safety Testing: Stress-testing LLMs to identify content moderation bypasses in Arabizi and regional dialects.
- RLHF & SFT Data Annotation: Rating, ranking, and refining Arabic AI responses for natural tone, grammar, and cultural accuracy.
- Custom Dataset Creation: Building tailored, high-quality parallel datasets (Arabizi ↔ Modern Standard Arabic ↔ English) on demand.
- Supervised Fine-Tuning (SFT) Data: Crafting custom prompt-and-response pairs to train AI on dialect-specific commands.
- Dialect & Arabizi Text Normalization: Creating custom text-cleansing mapping data to translate Arabizi/slang into Modern Standard Arabic (MSA) for clean AI pre-processing.
📩 Get in Touch for Enterprise & Freelance Projects:
- Email: Newmahod@gmail.com
- Buy Full Dataset on Gumroad: https://newmahod.gumroad.com/l/arabizi-dataset
- Downloads last month
- 28