Conversational Assistants to support Heart Failure Patients: comparing a Neurosymbolic Architecture with ChatGPT

  • 2025-04-24 18:16:24
  • Anuja Tayal, Devika Salunke, Barbara Di Eugenio, Paula Allen-Meares, Eulalia Puig Abril, Olga Garcia, Carolyn Dickens, Andrew Boyd
  • 0

Abstract

Conversational assistants are becoming more and more popular, including inhealthcare, partly because of the availability and capabilities of LargeLanguage Models. There is a need for controlled, probing evaluations with realstakeholders which can highlight advantages and disadvantages of moretraditional architectures and those based on generative AI. We present awithin-group user study to compare two versions of a conversational assistantthat allows heart failure patients to ask about salt content in food. Oneversion of the system was developed in-house with a neurosymbolic architecture,and one is based on ChatGPT. The evaluation shows that the in-house system ismore accurate, completes more tasks and is less verbose than the one based onChatGPT; on the other hand, the one based on ChatGPT makes fewer speech errorsand requires fewer clarifications to complete the task. Patients show nopreference for one over the other.

 

Quick Read (beta)

loading the full paper ...