Home // GLOBAL HEALTH 2023, The Twelfth International Conference on Global Health Challenges // View article
Authors:
Sareh Ahmadi
Edward A. Fox Ahmadi
Keywords: Episodic Future Thinking (EFT); delay discounting; maladaptive health behavior; Natural Language Processing (NLP); zero-shot learning; few-shot in-context learning.
Abstract:
Maladaptive health behaviors are closely linked to lifestyle-related diseases, such as obesity and type 2 diabetes. One significant factor contributing to maladaptive behavior is delay discounting, the tendency to prioritize immediate rewards over delayed ones. Episodic Future Thinking (EFT) is an intervention to reduce delay discounting and promote behavior change. EFT involves mentally simulating future events in a vivid manner, influencing decision-making and emotional well-being. Studies show EFT’s effectiveness in reducing delay discounting and its potential for improving various health behaviors, including exercise and medication adherence. However, EFT’s mechanisms of action and the conditions that impact its efficacy are unknown. This paper describes a study of EFT ‘cue texts’ to determine what makes them effective. It explains a new and efficient method to classify such texts with a few data, which can be used for further analysis to identify what characteristics of the texts lead to positive health outcomes. Classification framework is built using the FLAN-T5 large language model, with good results from zero-shot, and better results from few-shot in-context learning. This approach may be extended to address other behavioral health, wellness informatics, and technology-related approaches to global health challenges.
Pages: 32 to 37
Copyright: Copyright (c) IARIA, 2023
Publication date: September 25, 2023
Published in: conference
ISSN: 2308-4553
ISBN: 978-1-68558-112-1
Location: Porto, Portugal
Dates: from September 25, 2023 to September 29, 2023