Inclusive Easy-to-Read Generation for Individuals with Cognitive Impairments

2025/10/01 by Ledoyen, François, Dias, Gaël, Lechervy, Alexis +8
#Artificial Intelligence (cs.AI) #Computation and Language (cs.CL) #FOS: Computer and information sciences

paper · doi:10.48550/arxiv.2510.00691

Abstract

Ensuring accessibility for individuals with cognitive impairments is essential for autonomy, self-determination, and full citizenship. However, manual Easy-to-Read (ETR) text adaptations are slow, costly, and difficult to scale, limiting access to crucial information in healthcare, education, and civic life. AI-driven ETR generation offers a scalable solution but faces key challenges, including dataset scarcity, domain adaptation, and balancing lightweight learning of Large Language Models (LLMs). In this paper, we introduce ETR-fr, the first dataset for ETR text generation fully compliant with European ETR guidelines. We implement parameter-efficient fine-tuning on PLMs and LLMs to establish generative baselines. To ensure high-quality and accessible outputs, we introduce an evaluation framework based on automatic metrics supplemented by human assessments. The latter is conducted using a 36-question evaluation form that is aligned with the guidelines. Overall results show that PLMs perform comparably to LLMs and adapt effectively to out-of-domain texts.

Citations

• Exploring Large Language Models to generate Easy to Read content
• Automating Easy Read Text Segmentation
• Aging Up AAC: An Introspection on Augmentative and Alternative Communication Applications for Autistic Adults
• FinGPT: Large Generative Models for a Small Language
• Investigating Large Language Models and Control Mechanisms to Improve Text Readability of Biomedical Abstracts
• Platypus: Quick, Cheap, and Powerful Refinement of LLMs
• Overcoming Catastrophic Forgetting in Zero-Shot Cross-Lingual Generation
• Training language models to follow instructions with human feedback
• LaMDA: Language Models for Dialog Applications
• BARThez: a Skilled Pretrained French Sequence-to-Sequence Model
• Multilingual Denoising Pre-training for Neural Machine Translation
• BERTScore: Evaluating Text Generation with BERT
• Don't Give Me the Details, Just the Summary! Topic-Aware Convolutional\n Neural Networks for Extreme Summarization
• Attention Is All You Need
• Cost of disorders of the brain in Europe 2010
• Prefix-Tuning: Optimizing Continuous Prompts for Generation
• Prefix-Tuning: Optimizing Continuous Prompts for Generation

Related

• Accessible Text Tools for People with Cognitive Impairments and Non-Native Readers: Challenges and Opportunities
• Evaluating the Effectiveness of Direct Preference Optimization for Personalizing German Automatic Text Simplifications for Persons with Intellectual Disabilities
• Exploring Large Language Models to generate Easy to Read content
• CogBench: A Large Language Model Benchmark for Multilingual Speech-Based Cognitive Impairment Assessment
• Data and Approaches for German Text simplification -- towards an Accessibility-enhanced Communication