Arabic Question Generation (AQG) remains challenging due to the linguistic complexity of Arabic and the scarcity of large annotated resources. This paper presents an end-to-end answer-aware framework that transforms Arabic passages into coherent and contextually aligned questions. The model is an encoder–decoder architecture in which both the encoder and decoder are initialized from checkpoints pretrained on large-scale Arabic and multilingual text corpora. The encoder captures contextual semantic and syntactic representations of the passage and the answer span, while the decoder generates questions autoregressively, which are subsequently refined in a postprocessing module. To support systematic evaluation, a large collection of ~71k Arabic passage–question–answer triplets was compiled from multiple sources. Extensive experiments are conducted under a unified evaluation protocol across multiple pretrained encoder–decoder configurations. Our proposal achieved 20.75, 35.46, and 47.16 in BLEU, METEOR, and ROUGE-L outperforming baseline models. Finally, the obtained human evaluations score 4.42/5 assesses grammaticality, fluency, relevance, and answer correctness.
Lafkiar, Said, Alami, Hamza, En-nahnahi, Noureddine. From passages to questions: an end-to-end transformer-based framework for answer-aware Arabic question generation. // Information Technology and Management. – 2026. – DOI: 10.1007/s10799-026-00507-3. – URL: http://link.springer.com/openurl/fulltext?id=doi:10.1007/s10799-026-00507-3
Lafkiar, Said, Alami, Hamza, En-nahnahi, & Noureddine (2026). From passages to questions: an end-to-end transformer-based framework for answer-aware Arabic question generation. Information Technology and Management. https://doi.org/10.1007/s10799-026-00507-3
Lafkiar, et al. “From passages to questions: an end-to-end transformer-based framework for answer-aware Arabic question generation.” Information Technology and Management, 2026. doi:10.1007/s10799-026-00507-3.
@article{Lafkiar2026,
title = {From passages to questions: an end-to-end transformer-based framework for answer-aware Arabic question generation},
author = {Lafkiar and Said and Alami and Hamza and En-nahnahi and Noureddine},
journal = {Information Technology and Management},
year = {2026},
doi = {10.1007/s10799-026-00507-3},
url = {http://link.springer.com/openurl/fulltext?id=doi:10.1007/s10799-026-00507-3}
}
Competitive firms may share their digital resource to improve their efficiency in the industries of banks, insurance and hospitals. Their decisions about information security are important, but fail to receive enough attention. This...
User-generated content (UGC), such as health-related answers, is critical in online health communities, and financial incentives are a practical tool designed to motivate such content generation behaviors. Although the effect of...
The burgeoning trend of Short Video Live Streaming E-Commerce (SVLSE) prompts an exploration of factors impacting Consumer Purchase Intention. Integrating the S-O-R (Stimulus-Organism-Response) model with Affective Decision Theory, this...
This study aims to understand how to predict financial performance by systematically reviewing its conceptual evolution and integration with Artificial Intelligence (AI). This study analyses 85 articles from 74 Scopus-indexed journals...
Previous studies investigated manipulated reviews in a context where reviewers need to be validated purchasers, and posting fake reviews could be very costly. It was found that review manipulation is usually associated with low-quality...