1. MMC
  2. All Volumes and Issues
  3. Volume 11, Number 3, 2024
  4. An Arabic question generation system based on a shared BERT-base encoder-decoder architecture

An Arabic question generation system based on a shared BERT-base encoder-decoder architecture

MMC.
2024;
Volume 11, Number 3
: pp. 763–772
https://doi.org/10.23939/mmc2024.03.763
Received: January 21, 2024
Accepted: August 16, 2024

Lafkiar S., En Nahnahi N.  An Arabic question generation system based on a shared BERT-base encoder-decoder architecture.  Mathematical Modeling and Computing. Vol. 11, No. 3, pp. 763–772 (2024)

Authors:
  1. S. Lafkiar
  2. N. En Nahnahi
1
LISAC Laboratory, Faculty of Sciences Dhar El Mahraz, Sidi Mohamed Ben Abdellah University
2
LISAC Laboratory, Faculty of Sciences Dhar El Mahraz, Sidi Mohamed Ben Abdellah University

A Question Generation System (QGS) is a sophisticated piece of AI technology designed to automatically generate questions from a given text, document, or context.  Recently, this technology has gained significant attention in various fields, including education, and content creation.  As AI continues to evolve, these systems are likely to become even more advanced and viewed as an inherent part of any modern e-learning or knowledge assessment system.  In this research paper, we showcase the effectiveness of leveraging pre-trained checkpoints for Arabic questions generation.  We propose a Transformer-based sequence-to-sequence model that seamlessly integrates with publicly accessible pre-trained AraBERT checkpoints.  Our study focuses on evaluating the advantages of initializing our model, encompassing both the encoder and decoder, with these checkpoints.  As resources for Arabic language are still limited and the publicly datasets for question generation systems in Arabic are not available, we collected our dataset for this task from various existing question answering, we used this latter to train and test our model.  The experimental results show that our model yields performance was able to outperform existing Arabic question generation models in terms of the BLEU and METEOR scores,by achieving 20.29 as BLEU score and 30.73 for METEOR.  Finally, we assessed the capability of our model to generate contextually relevant questions.

question generation system
natural language processing
reading comprehension
transformers
transfer learning
  1. Rakangor S., Ghodasara Y.  Literature review of automatic question generation systems.  International Journal of Scientific and Research Publications.  5 (1), 1–5 (2015).
  2. Achtaich K., Achtaich N., Fagroud F. Z., Toumi H.  ALMA: Machine learning breastfeeding chatbot.  Mathematical Modeling and Computing.  10 (2), 487–497 (2023).
  3. Chali Y., Hasan S. A.  Towards Topic-to-Question Generation.  Computational Linguistics.  41 (1), 1–20 (2015).
  4. Yao X., Bouma G., Zhang Y.  Semantics-based Question Generation and Implementation.  Dialogue Discourse.  3 (2), 11–42 (2012).
  5. Bousmaha K. Z., Chergui N. H., Mbarek M. S. A., Belguith L. H.  AQG: Arabic Question Generator.  Revue d'Intelligence Artificielle.  34 (6), 721–729 (2020).
  6. Banou Z., Elfilali S., Benlahmar H.  Towards a polynomial approximation of support vector machine accuracy applied to Arabic tweet sentiment analysis.  Mathematical Modeling and Computing.  10 (2), 511–517 (2023).
  7. Lafkiar S., Hamza A., Zouitni M., Burmani N., Badir H., En Nahnahi N.  Attentive Neural Seq2Seq for Arabic Question Generation.  International Conference on Advanced Intelligent Systems for Sustainable Development.  802–816 (2022).
  8. Du X., Shao J., Cardie C.  Learning to Ask: Neural Question Generation for Reading Comprehension.  Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers).  1342–1352 (2017).
  9. El Moatez B. N., Elmadany A., Abdul-Mageed M.  AraT5: Text-to-text transformers for Arabic language generation.  Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers).  628–647 (2022).
  10. Alhashedi S., Suaib N. M., Bakri A.  Arabic Automatic Question Generation Using Transformer Model. EasyChair Preprint no. 8588 (2022).
  11. Kurdi G., Leo J., Parsia B., Sattler U., Al-Emari S.  A systematic review of automatic question generation for educational purposes.  International Journal of Artificial Intelligence in Education.  30, 121–204 (2020).
  12. Alazani S. A., Mahender C. N.  Rule based question generation for Arabic text: question answering system.  Proceedings of the International Conference on Data Science, Machine Learning and Artificial Intelligence.  7–12 (2021).
  13. Re O. C.  Building a system based on natural question generatio. Mohamed Elbasyouni (2014).
  14. Freydenberger D. D.  Extended regular expressions: Succinctness and decidability.  Theory of Computing Systems.  53, 159–193 (2013).
  15. Nagoudi E. M. B., Elmadany A., Abdul-Mageed M.  AraT5: Text-to-text transformers for Arabic language generation.  Preprint arXiv:2109.12068 (2021).
  16. Mozannar H., Maamary E., El Hajal K., Hajj H.  Neural Arabic Question Answering.  Proceedings of the Fourth Arabic Natural Language Processing Workshop.  108–118 (2019).
  17. Lewis P., Oğuz B., Rinott R., Riedel S., Schwenk H.  MLQA: Evaluating cross-lingual extractive question answering.  Preprint arXiv:1910.07475 (2019).
  18. Clark J. H., Choi E., Collins M., Garrette D., Kwiatkowski T., Nikolaev V., Palomaki J.  Tydi qa: A benchmark for information-seeking question answering in ty pologically di verse languages.  Transactions of the Association for Computational Linguistics.  8, 454–470 (2020).
  19. Vaswani A., Shazeer N., Parmar N., Uszkoreit J., Jones L., Gomez A. N., Kaiser Ł., Polosukhin I.  Attention Is All You Need.  Advances in Neural Information Processing Systems.  30, 1–11 (2017).
  20. Bas E.  A robust optimization approach to diet problem with overall glycemic load as objective function.  Applied Mathematical Modelling.  38 (19–20), 4926–4940 (2014).
  21. He K., Zhang X., Ren S., Sun J.  Deep residual learning for image recognition.  2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).  770–778 (2016).
  22. Ba J. L., Kiros J. R., Hinton G. E.  Layer normalization. Preprint arXiv:1607.06450 (2016).
  23. Devlin J., Chang M. W., Lee K., Toutanova K.  BERT: Pre-training of deep bidirectional transformers for language understanding.  Preprint arXiv:1810.04805 (2018).
  24. Rothe S., Narayan S., Severyn A.  Leveraging pre-trained checkpoints for sequence generation tasks.  Transactions of the Association for Computational Linguistics.  8, 264–280 (2020).
  25. Papineni K., Roukos S., Ward T., Zhu W.-J.  Bleu: a Method for Automatic Evaluation of Machine Translation.  Proceedings of the 40th Annual Meeting of the Association for Computational Linguistics.  311–318 (2002).
  26. Lavie A., Denkowski M. J.  The METEOR metric for automatic evaluation of machine translation.  Machine translation.  23 (2), 105–115 (2009).
  27. Lin C.-Y.  ROUGE: A Package for Automatic Evaluation of Summaries.  Proceedings of the Workshop on Text Summarization Branches Out.  74–81 (2004).
  28. Raffel C., Shazeer N., Roberts A., Lee K., Narang S., Matena M., Zhou Y., Li W., Liu P. J.  Exploring the limits of transfer learning with a unified text-to-text transformer.  The Journal of Machine Learning Research.  21 (1), 5485–5551 (2020).