Article Preview
TopIntroduction
Along with the global spread of COVID-19, people's demands for a healthy lifestyle have enhanced rapidly since 2019. Due to the recommendations of the World Health Organization (WHO) (Jones et al., 2020) on maintaining physical distance, healthcare-related services cannot provide face-to-face services. Countering this situation, several companies have been striving to provide mobile technology-based healthcare services (i.e., mobile healthcare service) to individuals who desire to take care of their health. Because mobile healthcare services do not require face-to-face communication, the services are suitable for the Covid-19 environment and have attracted a large number of users (Ming et al., 2020).
Mobile healthcare service is conceptualized as a system that offers health-related information and services via mobile communication technologies (e.g., smart phones and mobile networks) (Jaiswal and Anand, 2021; Zhao et al., 2018). Owing to the immense contribution to big data technology and artificial intelligence in recent times, mobile healthcare services can adequately trace the health condition of their users, and offer timely treatment and care (Zhao et al., 2018; Bol et al., 2018; Saheb, 2020). Mobile healthcare services not only save diagnosis time and cost but also play a positive role to improve the efficiency of medical resources (Lin et al., 2021). Therefore, a number of countries (e.g., Korea, China, and United States) have been taking efforts to support the development of mobile healthcare services (Zhao et al., 2018). Additionally, many people have been attracted by the advantages of mobile healthcare services and have resorted in their daily usage (Ming et al., 2020). As of 2021, more than 400,000 mobile healthcare applications have been made available on the Google play store (Imaginovation, 2021). According to Global market insights (2021), the global mobile health market has been estimated to reach $289.4 billion by 2025.
Considering user satisfaction as being related to user loyalty and the intention to use specific technologies continuously (Bhattacherjee, 2001; Kumar et al., 2013; Cheng and Jiang, 2020), several studies have explored the user experience elements that lead to greater user satisfaction with mobile healthcare services (e.g., Oppong et al., 2021; Handayani et al., 2018; Keikhosrokiani et al., 2020). However, most studies exploring the user experience of mobile healthcare services only employed a small number of samples and traditional methodologies (e.g., analyzing small survey-based data with structural equation model). Therefore, to solve these shortcomings, this research attempts to predict and explore user satisfaction with mobile healthcare services by examining big data using machine learning approaches.
Specifically, 139,604 usable online reviews of a particular mobile healthcare service (i.e., Samsung Health) from Google Play store were collected. Subsequently, five machine learning classifiers (i.e., Logistic regression, Random Forest, Gradient Boosting Model, Extreme Gradient Boosting, Naïve Bayes) with three word embedding methods (Bag-of-words (BoW), Term Frequency - Inverse Document Frequency (TF-IDF), Global Vectors for Word Representation (GloVe)) for predicting user satisfaction with healthcare services were applied. It is worth noting that, this is one of the first studies to predict user satisfaction by employing big data and machine learning approaches in the mobile healthcare service field. Theoretically, this study also can contribute to expanding the literature of affect theory (Kratzwald et al., 2018) which indicates that user-generated contents are notably related to user satisfaction with certain services and products.