Modelling and predicting User Engagement in mobile applications


The mobile ecosystem is dramatically growing towards an unprecedented scale, with an extremely crowded marketand fierce competition among app developers. Today, keeping users engaged with a mobile app is key for its success since userscan remain active consumers of services and/or producers of new contents. However, users may abandon a mobile app at anytime due to various reasons,e.g.,the success of competing apps, decrease of interest in the provided services, etc. In this context,predicting when a user may get disengaged from an app is an invaluable resource for developers, creating the opportunity toapply intervention strategies aiming at recovering from disengagement (e.g.,sending push notifications with new contents).In this study, we aim at providing evidence that predicting when mobile app users get disengaged is possible with a good levelof accuracy. Specifically, we propose, apply, and evaluate a framework to model and predict User Engagement (UE) in mobileapplications via different numerical models. The proposed framework is composed of an optimized agglomerative hierarchicalclustering model coupled to (i) a Cox proportional hazards, (ii) a negative binomial, (iii) a random forest, and (iv) a boosted-treemodel.The proposed framework is empirically validated by means of a year-long observational dataset collected from a real deployment of a waste recycling app. Our results show thatin this contextthe optimized clustering model classifies users adequatelyand improves UE predictability for all numerical models. Also, the highest levels of prediction accuracy and robustness areobtained by applying either the random forest classifier or the boosted-tree algorithm.

Data Science