TY - JOUR
T1 - A text style transfer system for reducing the physician–patient expertise gap
T2 - An analysis with automatic and human evaluations
AU - Bacco, Luca
AU - Dell'Orletta, Felice
AU - Lai, Huiyuan
AU - Merone, Mario
AU - Nissim, Malvina
N1 - Funding Information:
The authors would like to express their sincere gratitude to the physicians of the Department of Orthopaedic Surgery at the University Campus Bio-Medico of Rome, Italy, for their professional contribution to the development of this work. Special thanks are due to L. Ambrosio (MD), G. Papalia (MD), F. Russo (MD), and G. Vadalà (MD) for their invaluable assistance. The authors would also like to acknowledge the efforts and feedback of all the experts and lay annotators, too, who played a crucial role in the study. Their contribution is greatly appreciated.
Publisher Copyright:
© 2023 Elsevier Ltd
PY - 2023/12/15
Y1 - 2023/12/15
N2 - Physicians and patients often come from different backgrounds and have varying levels of education, which can result in communication difficulties in the healthcare process. To address this expertise gap, we present a “Text Style Transfer” system. Our system uses Semantic Textual Similarity techniques based on Sentence Transformers models to create pseudo-parallel datasets from a large, non-parallel corpus of lay and expert texts. This approach allowed us to train a denoising autoencoder model (BART), overcoming the limitations of previous systems. Our extensive analysis, which includes both automatic metrics and human evaluations from both lay (patients) and expert (physicians) individuals, shows that our system outperforms state-of-the-art models and is comparable to human-provided gold references in some cases.
AB - Physicians and patients often come from different backgrounds and have varying levels of education, which can result in communication difficulties in the healthcare process. To address this expertise gap, we present a “Text Style Transfer” system. Our system uses Semantic Textual Similarity techniques based on Sentence Transformers models to create pseudo-parallel datasets from a large, non-parallel corpus of lay and expert texts. This approach allowed us to train a denoising autoencoder model (BART), overcoming the limitations of previous systems. Our extensive analysis, which includes both automatic metrics and human evaluations from both lay (patients) and expert (physicians) individuals, shows that our system outperforms state-of-the-art models and is comparable to human-provided gold references in some cases.
KW - Healthcare
KW - Natural language processing
KW - Semantic textual similarity
KW - Text simplification
KW - Text style transfer
UR - http://www.scopus.com/inward/record.url?scp=85165158406&partnerID=8YFLogxK
U2 - 10.1016/j.eswa.2023.120874
DO - 10.1016/j.eswa.2023.120874
M3 - Article
AN - SCOPUS:85165158406
SN - 0957-4174
VL - 233
JO - Expert systems with applications
JF - Expert systems with applications
M1 - 120874
ER -