Paper ID
2328
Paper Type
short
Description
Cancer patients’ emotions are important to their health and outcomes. In light of appraisal theory and optimal matching theory, a viable way to create positive changes in patient emotion is through the match between their indicated needs and provided support, as well as the relevance of the support. In this study, we extend extant literature by empirically examining the effects of matched/unmatched patient need and relevant social support on their emotion change. We develop a novel deep learning based multi-label classification method to categorize different types of patient needs and social support. Our proposed method surpasses seven robust benchmark models. The empirical analysis suggests relevant support can increase patient happiness while unmatched patient needs negatively impact happiness. Both the matched needs and relevant support are associated with increased fear, but this effect can be reduced by including more images in replies. This study provides several implications to community operators or facilitators on how to encourage members’ participation and to provide the needed support to patients.
Recommended Citation
Xu, Anqi; Liu, Xiao; and Hu, Paul Jen-Hwa, "Examining Effects of Fit between Patient Need and Social Support: A Deep Learning Based Multi-Label Classification Approach" (2019). ICIS 2019 Proceedings. 13.
https://aisel.aisnet.org/icis2019/is_health/is_health/13
Examining Effects of Fit between Patient Need and Social Support: A Deep Learning Based Multi-Label Classification Approach
Cancer patients’ emotions are important to their health and outcomes. In light of appraisal theory and optimal matching theory, a viable way to create positive changes in patient emotion is through the match between their indicated needs and provided support, as well as the relevance of the support. In this study, we extend extant literature by empirically examining the effects of matched/unmatched patient need and relevant social support on their emotion change. We develop a novel deep learning based multi-label classification method to categorize different types of patient needs and social support. Our proposed method surpasses seven robust benchmark models. The empirical analysis suggests relevant support can increase patient happiness while unmatched patient needs negatively impact happiness. Both the matched needs and relevant support are associated with increased fear, but this effect can be reduced by including more images in replies. This study provides several implications to community operators or facilitators on how to encourage members’ participation and to provide the needed support to patients.