Currently submitted to: JMIR mHealth and uHealth
Date Submitted: Jan 7, 2025
Open Peer Review Period: Feb 18, 2025 - Apr 15, 2025
(currently open for review)
Warning: This is an author submission that is not peer-reviewed or edited. Preprints - unless they show as "accepted" - should not be relied on to guide clinical practice or health-related behavior and should not be reported in news media as established information.
Cluster-Based Predictive Modelling of User Ratings for Physical Activity Apps Using MARS Dimensions
ABSTRACT
Background:
The rapid proliferation of mobile health applications has underscored the need for robust tools to evaluate their quality. The Mobile App Rating Scale (MARS) is a widely used framework to assess key app dimensions—Engagement, Functionality, Aesthetics, and Information—that influence user satisfaction. Despite its utility, limited research has leveraged MARS dimensions to predict user ratings effectively.
Objective:
This study investigates how K Means clustering, combined with machine learning models, can predict user ratings for physical activity apps based on MARS dimensions, with the goal of forecasting ratings prior to production and uncovering insights into user satisfaction drivers.
Methods:
The dataset comprises MARS-rated physical activity apps with user ratings. Initially, exploratory data analysis provided insights into generalized patterns among MARS dimensions. K Means clustering segmented the apps into two clusters, enhancing data homogeneity. Predictive models were then applied to both the unclustered dataset and the clustered subsets, with performance evaluated based on accuracy, mean absolute error (MAE), and R-squared values. The top-performing model from each cluster was selected for combined analysis to further improve prediction reliability.
Results:
In the unclustered dataset, KNN achieved the highest accuracy (86.36%) with an MAE of 0.26 and R-squared of 0.06. Post-clustering, SVR achieved the highest accuracy in Cluster 1 (88.89%, MAE 0.27, R-squared 0.05), and KNN excelled in Cluster 2 (88.24%, MAE 0.28, R-squared 0.03). Combining the predictions of these models for the two clusters improved overall accuracy to 88.64%, with MAE of 0.27 and R-squared of 0.04.
Conclusions:
The combined clustering and model approach enhances prediction accuracy and reveals how user satisfaction drivers vary across app types. This tailored methodology offers insights for app developers aiming to optimize apps, aligning features with distinct user needs for higher satisfaction and engagement.
Citation
Request queued. Please wait while the file is being generated. It may take some time.
Copyright
© The authors. All rights reserved. This is a privileged document currently under peer-review/community review (or an accepted/rejected manuscript). Authors have provided JMIR Publications with an exclusive license to publish this preprint on it's website for review and ahead-of-print citation purposes only. While the final peer-reviewed paper may be licensed under a cc-by license on publication, at this stage authors and publisher expressively prohibit redistribution of this draft paper other than for review purposes.