Datasets:
metadata
license: cc-by-4.0
language:
- ar
- en
task_categories:
- translation
- text-generation
pretty_name: Arabic-English Sentence Bank 25k
size_categories:
- 10K<n<100K
tags:
- translation
- nlp
- text
Arabic-English Sentence Bank 25k
Arabic-English Sentence Bank 25k is a diverse bilingual dataset designed to support Arabic-English natural language processing tasks. It contains 25,000 sentence pairs in Arabic and English, covering a wide variety of sentence types, styles, and communicative functions. The dataset was created with the assistance of artificial intelligence and curated to provide broad coverage of common language patterns.
Dataset Size
The dataset contains 25,000 Arabic-English sentence pairs
Sentence Categories
The dataset consists of the following sentence types:
| Category | Number of Sentences |
|---|---|
| Inspirational Sentences | 2,500 |
| Factual Sentences | 2,000 |
| Opinions | 2,000 |
| Recommendations | 2,000 |
| Questions | 2,000 |
| Hypothetical Sentences | 2,000 |
| Exclamatory Sentences | 2,000 |
| Cause-Effect Sentences | 1,500 |
| Conditional Sentences | 1,500 |
| Commands / Requests | 1,500 |
| Everyday Expressions | 1,000 |
| Contrasting Sentences | 1,000 |
| Time and Quantity Sentences | 1,000 |
| Explanations | 1,000 |
| Negative Sentences | 1,000 |
| Comparative Sentences | 1,000 |
Dataset Columns
The recommended structure of the dataset is:
| Column | Description |
|---|---|
Arabic |
Arabic sentence |
English |
English version of the sentence |
Type |
Sentence category or type |
Example:
| Arabic | English | Type |
|---|---|---|
| النجاح يبدأ بخطوة صغيرة. | Success begins with a small step. | Inspirational Sentences |
| هل يمكنك مساعدتي في هذا المشروع؟ | Can you help me with this project? | Questions |
Citation
@dataset{alselwi_2025_arabic_english_sentence_bank_25k,
author = {Taha Alselwi},
title = {Arabic-English Sentence Bank 25k},
year = {2025},
}