Using machine learning to mine user reviews from hypertension management mHealth Apps to explore user satisfaction influencing factors and their asymmetry: A comparative study (Preprint)

Author:

He YunfanORCID,Zhu Wei,Wang TongORCID,Chen Han,Xin Junyi,Liu Yongcheng,Lei JianboORCID,Liang JunORCID

Abstract

BACKGROUND

The use of hypertension management applications (HMAs) remains unsatisfactory. Currently, there is a lack of real-world research based on big data and exploratory mining comparing Chinese and American HMAs.

OBJECTIVE

To use data mining to compare HMA user experience, satisfaction level, influencing factors, and asymmetry for Chinese, and American users; evaluate the differences between satisfaction and its influencing factors; and explore the asymmetry of the factors.

METHODS

HMAs and user reviews were obtained from 10 major Chinese and American App stores worldwide. The latent dirichlet allocation topic model identified user review topics. The Tobit model was used to explore the effects and differences of each topic on user satisfaction. The Wald test was used to analyze effect differences.

RESULTS

We included 261 HMAs with user reviews and 116,686 user reviews. Chinese HMAs (91 vs. 220) and reviews (16,561 vs. 100,125) were fewer than their American counterparts. The overall HMA user satisfaction rate was 75.22%, with a higher satisfaction with Chinese HMAs (83.73% vs. 73.81%). Eight factors significantly affected the positive rating deviation (PD) of Chinese HMA user satisfaction, and nine factors for the negative rating deviation. All 12 factors significantly affected the PD and ND of American HMA user satisfaction. The effects of Chinese and American HMA user satisfaction factors significantly differed in the positive deviation and negative.

CONCLUSIONS

User satisfaction factors in different countries were asymmetric and considerably different. Cost, measurement accuracy, and compatibility mainly affected Chinese HMA user dissatisfaction. Data sharing, synchronization, software reliability, compatibility, and advertisement distribution were basically required by American users. Personalized experience plans based on user groups should be developed in different countries.

Publisher

JMIR Publications Inc.

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3