saidlafkiar82/ARQGData

ARQGData Corpus

Dataset Description

This repository contains the complete ARQGData Corpus, a curated dataset designed for Arabic Automatic Question Generation (AQG). It is intended for use in training, testing, and evaluation of deep learning models. The dataset provides high-quality, diverse examples to facilitate research and development in Arabic natural language processing and educational technology applications.

• —Language: Arabic
• —Dataset Type: CSV
• —Size: 71148

Data Fields

Each row in the CSV file contains the following fields:

• —Contexte: Text passage from which a question can be generated
• —question: The generated question corresponding to the context
• —answer: The answer extracted from the context and used as the target answer for the question.
• —answer_start: Character-level starting position of the answer within the context. For instances containing multiple answers, the corresponding start positions are separated using the | symbol.

Data Splits

• —Train: 56920
• —Validation: 7114
• —Test: 7114

Usage

You can easily load the dataset using the Hugging Face datasets library:

from datasets import load_dataset

dataset = load_dataset("username/ARQGData-Corpus", split="train")
print(dataset[0])


---
# License
This dataset is released under the CC BY 4.0 license.
Free to use, share, and modify for non-commercial research and educational purposes,
with proper attribution and distribution of derivative works under the same license.

#Citation

If you use this dataset in your work, please cite it as follows:
@dataset{lafkiar2025arqgdata,
  title={ARQGData Corpus},
  author={Said Lafkiar, Hamza Alami, and Noureddine En-nahnahi},
  year={2025},
  url={https://huggingface.co/datasets/saidlafkiar82/ARQGData}
}