Machine learning for mHealth apps quality evaluation: An approach based on user feedback analysis

Mariem Haoues; Raouia Mokni; Asma Sellami

doi:10.1007/s11219-023-09630-8

Machine learning for mHealth apps quality evaluation: An approach based on user feedback analysis

Softw Qual J. 2023 May 23:1-31. doi: 10.1007/s11219-023-09630-8. Online ahead of print.

Authors

Mariem Haoues^{1

2}, Raouia Mokni^{3

4}, Asma Sellami²

Affiliations

¹ Department of Software Engineering, College of Computer Engineering and Sciences, Prince Sattam bin Abdulaziz University Al-Kharj, Al-Kharj, 11942 Saudi Arabia.
² Mir@cl Laboratory, University of Sfax, ISIMS, BP 242. 3021., Sfax, Tunisia.
³ Department of Information System, College of Computer Engineering and Sciences, Prince Sattam bin Abdulaziz University Al-Kharj, Al-Kharj, 11942 Saudi Arabia.
⁴ University of Sfax, Sfax, Tunisia.

Abstract

Mobile apps for healthcare (mHealth apps for short) have been increasingly adapted to help users manage their health or to get healthcare services. User feedback analysis is a pertinent method that can be used to improve the quality of mHealth apps. The objective of this paper is to use supervised machine learning algorithms to evaluate the quality of mHealth apps according to the ISO/IEC 25010 quality model based on user feedback. For this purpose, a total of 1682 user reviews have been collected from 86 mHealth apps provided by Google Play Store. Those reviews have been classified initially into the ISO/IEC 25010 eight quality characteristics, and further into Negative, Positive, and Neutral opinions. This analysis has been performed using machine learning and natural language processing techniques. The best performances were provided by the Stochastic Gradient Descent (SGD) classifier with an accuracy of 82.00% in classifying user reviews according to the ISO/IEC 25010 quality characteristics. Moreover, Support Vector Machine (SVM) classified the collected user reviews into Negative, Positive, and Neutral with an accuracy of 90.50%. Finally, for each quality characteristic, we classified the collected reviews according to the sentiment polarity. The best performance results were obtained for the Usability, Security, and Compatibility quality characteristics using SGD classifier with an accuracy equal to 98.00%, 97.50%, and 96.00%, respectively. The results of this paper will be effective to assist developers in improving the quality of mHealth apps.

Keywords: ISO/IEC 25010 quality model; Machine learning; Natural language processing; Quality characteristics; Sentiment analysis; User feedback; mHealth apps.

© The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2023, Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.