Spoiled for Choice? Personalized Recommendation for Healthcare Decisions: A Multi-Armed Bandit Approach

39 Pages Posted: 13 Jun 2019 Last revised: 10 Oct 2021

See all articles by Tongxin Zhou

Tongxin Zhou

Arizona State University (ASU) - Department of Information Systems

Yingfei Wang

University of Washington - Michael G. Foster School of Business

Lu (Lucy) Yan

Indiana University - Kelley School of Business - Department of Operation & Decision Technologies

Yong Tan

University of Washington - Michael G. Foster School of Business

Date Written: September 13, 2020

Abstract

Social media-based platforms tend to provide much information and numerous options for users, making choice overload a prevalent issue in many online communities. Online healthcare communities (OHCs), which provide users with various healthcare interventions to promote healthy behavior and improve adherence, are no exception. When faced with too many choices, however, individuals may find it difficult to decide which behavior intervention to take, especially when they lack the experience or knowledge to evaluate different options. Consequently, the choice overload issue may negatively affect users’ engagement in health management in OHCs. In this study, we take a design-science perspective to propose a recommendation framework that helps users to select healthcare interventions. Taking into account that users’ health behaviors can be highly dynamic and diverse, we propose a multi-armed bandit (MAB)-driven recommendation framework, which enables us to adaptively learn users’ preference variations while promoting recommendation diversity in the meantime. To better adapt an MAB to the healthcare context, we synthesize two innovative model components based on prominent health theories. The first component is a deep-learning-based feature engineering procedure, which is designed to learn crucial recommendation contexts in regard to users’ sequential health histories, health-management experiences, preferences, and intrinsic attributes of healthcare interventions. The second component is a diversity constraint, which structurally diversifies recommendations in different dimensions to provide well-rounded support for individuals’ health management. We apply our approach to an online weight management context and evaluate it rigorously through a series of experiments. Our results demonstrate that each of the design components is effective and that our recommendation design outperforms a wide range of state-of-the-art recommendation systems. Our study contributes to the emerging IS research on the application of business intelligence. The results have important implications for multiple stakeholders, including online healthcare platforms, policymakers, and users.

Keywords: personalized healthcare recommendations, health behavior dynamics, recommendation diversity, multi-armed bandit (MAB), deep-learning embeddings

Suggested Citation

Zhou, Tongxin and Wang, Yingfei and Yan, Lu (Lucy) and Tan, Yong, Spoiled for Choice? Personalized Recommendation for Healthcare Decisions: A Multi-Armed Bandit Approach (September 13, 2020). Kelley School of Business Research Paper No. 19-25, Available at SSRN: https://ssrn.com/abstract=3397452 or http://dx.doi.org/10.2139/ssrn.3397452

Tongxin Zhou (Contact Author)

Arizona State University (ASU) - Department of Information Systems ( email )

Tempe, AZ
United States

Yingfei Wang

University of Washington - Michael G. Foster School of Business ( email )

Box 353200
Seattle, WA 98195-3200
United States

Lu (Lucy) Yan

Indiana University - Kelley School of Business - Department of Operation & Decision Technologies ( email )

Department of Operations and Decision Technologies
1309 E. Tenth Street
Bloomington, IN 47401
United States

Yong Tan

University of Washington - Michael G. Foster School of Business ( email )

Box 353226
Seattle, WA 98195-3226
United States

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Downloads
229
Abstract Views
1,115
rank
184,526
PlumX Metrics