ORCID
- Jonathan Turner: 0000-0001-9991-3130
- Damon Berry: 0000-0002-2290-7661
- Dympna O'Sullivan: 0000-0003-2841-9738
Abstract
There is much information available on dementia, but information may not be written at an appropriate level for a lay reader. One aim of the CoDESIGN project is to simplify information available for those who have lower levels of literacy. We investigate the potential to simplify documents to reflect the literacy abilities of the reader. We report on the ability of ChatGPT to modify documents to such target reading levels while retaining the accuracy and completeness of the documents, while avoiding the introduction of errors. Results suggest that human supervision is required before documents can be made available to readers.
Keywords
ChatGPT, Dementia, Large Language Models, literacy, readability
DOI Link
Publication Date
2025-05-15
Publication Title
Studies in Health Technology and Informatics
Volume
327
First Page
880
Last Page
881
Deposit Date
2025-07-23
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 International License.
Additional Links
Recommended Citation
Turner, Jonathan; Berry, Damon; Doyle, Julie; Kambampati, Harish; Moran, Orla; Smith, Suzanne; Wilson, Michael; and O'Sullivan, Dympna, "Improving the Readability of Dementia Information Documents Using ChatGPT" (2025). Research Outputs: 2025-Present. 2.
https://arrow.tudublin.ie/teapotro/2