This paper presents the first Arabic crossword puzzle generator driven by advanced AI tech- nology. Leveraging cutting-edge large lan- guage models including GPT4, GPT3-Davinci, GPT3-Curie, GPT3-Babbage, GPT3-Ada, and BERT, the system generates distinctive and challenging clues. Based on a dataset compris- ing over 50,000 clue-answer pairs, the genera- tor employs fine-tuning, few/zero-shot learning strategies, and rigorous quality-checking proto- cols to enforce the generation of high-quality clue-answer pairs. Importantly, educational crosswords contribute to enhancing memory, expanding vocabulary, and promoting problem- solving skills, thereby augmenting the learn- ing experience through a fun and engaging ap- proach, reshaping the landscape of traditional learning methods. The overall system can be exploited as a powerful educational tool that amalgamates AI and innovative learning tech- niques, heralding a transformative era for Ara- bic crossword puzzles and the intersection of technology and education.

Zeinalipour, K., Saad, M., Maggini, M., Gori, M. (2023). ArabIcros: AI-Powered Arabic Crossword Puzzle Generation for Educational Applications. In Proceedings of ArabicNLP 2023 (pp.288-301). Association for Computational Linguistics [10.18653/v1/2023.arabicnlp-1.23].

ArabIcros: AI-Powered Arabic Crossword Puzzle Generation for Educational Applications

Kamyar Zeinalipour;Mohamed Saad;Marco Maggini;Marco Gori
2023-01-01

Abstract

This paper presents the first Arabic crossword puzzle generator driven by advanced AI tech- nology. Leveraging cutting-edge large lan- guage models including GPT4, GPT3-Davinci, GPT3-Curie, GPT3-Babbage, GPT3-Ada, and BERT, the system generates distinctive and challenging clues. Based on a dataset compris- ing over 50,000 clue-answer pairs, the genera- tor employs fine-tuning, few/zero-shot learning strategies, and rigorous quality-checking proto- cols to enforce the generation of high-quality clue-answer pairs. Importantly, educational crosswords contribute to enhancing memory, expanding vocabulary, and promoting problem- solving skills, thereby augmenting the learn- ing experience through a fun and engaging ap- proach, reshaping the landscape of traditional learning methods. The overall system can be exploited as a powerful educational tool that amalgamates AI and innovative learning tech- niques, heralding a transformative era for Ara- bic crossword puzzles and the intersection of technology and education.
2023
9781959429272
Zeinalipour, K., Saad, M., Maggini, M., Gori, M. (2023). ArabIcros: AI-Powered Arabic Crossword Puzzle Generation for Educational Applications. In Proceedings of ArabicNLP 2023 (pp.288-301). Association for Computational Linguistics [10.18653/v1/2023.arabicnlp-1.23].
File in questo prodotto:
File Dimensione Formato  
2023.arabicnlp-1.23.pdf

accesso aperto

Tipologia: PDF editoriale
Licenza: PUBBLICO - Pubblico con Copyright
Dimensione 3.95 MB
Formato Adobe PDF
3.95 MB Adobe PDF Visualizza/Apri

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11365/1253155