Understanding Mobile App Usage through Contextual Data: An Analysis Using Linear Regression

Abstract

The use of mobile apps leaped into the scene over the past few years, and the issue of features contributing to app engagement is vital to developers, marketers, and researchers. In this paper, the researcher attempts to compare the pattern of mobile application usage based on contextual information through the application of the linear regression method using parameters considered important to determine the location, time of day, and the type of device used. We have already seen a trend of research showing how contextual information can influence the behavior of the user (Zhao et al., 2019), although the challenge remains to understand how those factors combine to impact the usage of the apps on a wider scale. Our approach is to use linear regression models to forecast app engagement on the basis of the contextual features by use of data set encompassing both the time-series data of app usages as well as the context of a user. We find that time of day along with user location has a significant impact on mobile app usage with time having the biggest impact. The results have been uploaded to the existing literature in the field of mobile computing and app personalization, providing an understanding of how to enhance the app design and the experience with encountering it. The article under consideration underlines the necessity of the implementation of real-time contextual data in the process of boosting app utilization and customer satisfaction.