HHBBHHHHBV
A Multimodal Intelligent approach for
Dyslexia Detection and Reading Support
Mehdi Belguith¹, Wafa Khabou¹, Latifa Iben Nasr¹
¹ANLP-RG, MIRACL Lab. FSEGS, University of Sfax, Tunisia
Email: belguith.mehdi2017@gmail.com, latifa.ibennasr@fsegs.usf.tn, waafakhabou@gmail.com
The 4th International Conference
on Language Processing
and Knowledge Management
Data collection
For dyslexia detection:
Available multimodal datasets
For dyslexia reading support:
Construction of a specific dataset based on Tunisian textbooks
Proposed models
Dyslexia Detection:
Hybrid method for multimodal data
Dyslexia-ready support:
LLMs, NLP techniques, and ML models for error detection and individualized support
This research proposes a multimodal approach (text, speech, eye-tracking, images, and videos) for:
The proposed method for dyslexia support is based on NLP and machine learning techniques.
As perspectives, we intend to apply and evaluate eye-tracking models to detect dyslexia in children.
[1] Olusanya BO, Smythe T, Ogbo FA, Nair MKC, Scher M, Davis AC. Global prevalence of developmental disabilities in children and adolescents: A systematic umbrella review. Front Public Health. 2023 Feb 16;11:1122009. doi: 10.3389/fpubh.2023.1122009. PMID: 36891340; PMCID: PMC9987263.
[2] Martín-Ruiz, I., Rueda-Flores, E., Infante-Cañete, L., Alarcón-Orozco, E., & Robles-Sánchez, M. J. (2026). Identification and Detection of Specific Learning Disabilities: A Systematic Review. Education Sciences