User Engagement in Mobile Applications: Modeling and Predicting
Keywords:
User Engagement, mobile apps, numerical modelling, clusteringAbstract
The mobile ecosystem is expanding rapidly, making app developers highly competitive. To keep users engaged, apps can anticipate when they might disengage, allowing developers to implement intervention tactics. This work demonstrates that it is possible to accurately forecast user disengagement using numerical models. The framework includes Cox proportional hazards, negative binomial, random forest, boosted-tree, and optimized agglomerative hierarchical clustering models. A 12-month observation dataset from a trash recycling app development validates the approach. The optimal clustering model enhances UE predictability and user identification, with random forest classifiers or boosted-tree methods yielding the best results.
References
Android. (2017). Android developer portal. [Online]. http://developer.android.com/about/index.html.
Liaw A, Wiener M. Classification and regression by Random Forest. R News. 2002; 2(3): 18–22. http://CRAN.R-project.org/doc/Rnews/.
Racherla P, Furner C, Babb J. Conceptualizing the implications of mobile app usage and stickiness: A research agenda. SSRN Electronic Journal. 2012. http://ssrn.com/abstract=2187056.
Anderlucci L. Comparing different approaches for clustering categorical data. PhD thesis. University of Bologna; 2012. http://amsdottorato.unibo.it/id/eprint/4302.
van den Hoven J. Clustering with optimized weights for Gower’s metric. Master’s thesis. Amsterdam, The Netherlands: Vrij University; 2016. https://science.vu.nl/en/Images/stageverslag-hoven_tcm296-777817.pdf.
Statista. (2018). Number of available applications in the Google Play Store from December 2009 to June 2018. [Online]. https://www.statista.com/statistics/266210/number-of-available-in-the-google-play-store/.
Chen T, He T. Higgs boson discovery with boosted trees. In: NIPS 2014 Workshop on High-Energy Physics and Machine Learning. 2015; 42: 69–80. http://proceedings.mlr.press/v4/chen14.html.
Lella A, Lipsman A. (2016). The 2016 U.S. mobile app report. comsCore white paper. https://www.comscore.com/Insights/Presentations-and-Whitepaperrs/2016/The-2016-US-Mobile-App-Report.
Attfield S, Lalmas M, Piwowarski B. Towards a science of user engagement (Position Paper). In: WSDM Workshop on User Modelling for Web Applications. 2011. https://eprints.mdx.ac.uk/id/eprint/8642.
Ramesh A, Goldwasser D, Huang B, Daume H III, Getoor L. Learning latent engagement patterns of students in online co urses. In: AAAI Conference on Artificial Intelligence. 2014; 28(1): 8920. https://dl.acm.org/citation.cfm?id=2894071
