Iterative Prompt Refinement for Dyslexia-Friendly Text Summarization Using GPT-4o

2026-02-25

AI Summary

This study addresses the challenges individuals with dyslexia face in reading fluency and comprehension due to textual complexity by proposing an iterative prompt optimization pipeline based on GPT-4o. The framework automatically generates simplified summaries that balance high readability with semantic fidelity through multiple rounds of generation and evaluation. This work introduces, for the first time, an iterative prompt refinement mechanism tailored specifically for text simplification targeting readers with dyslexia and establishes the first quantifiable benchmark for accessible summarization. Experimental results on a corpus of 2,000 news articles demonstrate that the majority of generated summaries achieve a Flesch Reading Ease score of at least 90 within four iterations, while maintaining a stable trade-off between readability and semantic preservation at approximately 0.55.

Technology Category

• Natural Language Processing: Summarization
• Humans and AI: AI for Accessibility
• Search and Optimization: Distributed Search

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Web Mining and Content Analysis: Normalization, clustering, classification, and summarization of Web text
• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work

Abstract

Dyslexia affects approximately 10% of the global population and presents persistent challenges in reading fluency and text comprehension. While existing assistive technologies address visual presentation, linguistic complexity remains a substantial barrier to equitable access. This paper presents an empirical study on dyslexia-friendly text summarization using an iterative prompt-based refinement pipeline built on GPT-4o. We evaluate the pipeline on approximately 2,000 news article samples, applying a readability target of Flesch Reading Ease >= 90. Results show that the majority of summaries meet the readability threshold within four attempts, with many succeeding on the first try. A composite score combining readability and semantic fidelity shows stable performance across the dataset, ranging from 0.13 to 0.73 with a typical value near 0.55. These findings establish an empirical baseline for accessibility-driven NLP summarization and motivate further human-centered evaluation with dyslexic readers.

Problem

• dyslexia
• text summarization
• readability
• accessibility
• linguistic complexity

Innovation

• iterative prompt refinement
• dyslexia-friendly summarization
• readability optimization
• GPT-4o
• accessibility-driven NLP