Text Simplification for Accessibility

Abstract

This project approached text simplification as an accessibility problem, not just a generation task, with emphasis on preserving meaning while reducing reading difficulty.

The work framed simplification quality around reader usefulness, grammatical clarity, and fidelity to the original text rather than model novelty alone.

What I Built

• Text simplification is not just an NLP generation task; usefulness depends on preserving meaning while reducing reading difficulty.
• The project's strongest framing is accessibility rather than model novelty.

Impact

• Connected NLP work to reader-facing accessibility outcomes instead of treating simplification as generic text generation.
• Emphasized evaluation criteria that matter for real users, including clarity, meaning preservation, and reading effort.