Abstract
Next app prediction can help enhance user interface design, pre-loading of apps, and network optimizations. Prior work has explored this topic, utilizing multiple different approaches but challenges like the user cold-start problem, data sparsity, and privacy concerns related to contextual data like location histories, persist. The user cold-start problem occurs when a user has recently registered to the smartphone app system and there is not enough information about his/her preferences and his/her history of smartphone usage. In this work, we try to address the above issues. We introduce WhatsNextApp, an approach based on LSTM (Long Short-Term Memory) networks using sequences of app usage logs. Our approach is inspired by Word Embeddings and treats sequences of app usage logs as sequences of words. We collect a real-life data set consisting of 975 Android users with over 22 million app usage events. We build a generic (user-independent) WhatsNextApp model and the evaluation with our data set shows that it outperforms related studies for existing users where we achieve a recall@8 (recall for the top 8 apps) of 92%. For the user cold-start problem with the 500 most frequent apps, we achieve a recall@8 of 82.7%.
Highlights
E VERY day, smartphone users use a variety of apps for different purposes [1]
We see three main benefits from predicting app usage: user interface optimization for enhancing usability and accessibility, loading apps that are about to be used into memory or keeping them in memory, and optimizing network infrastructure, as different apps produce different amounts of traffic with different endpoints
We propose the WhatsNextApp approach, a deep learning approach based on Long Short-Term Memory Networks (LSTMs) that considers the sequences of app usage and is inspired by the concept of word embeddings
Summary
E VERY day, smartphone users use a variety of apps for different purposes [1]. The number of apps in Google Play and the Apple App Store already exceeds 2.8 million apps and 2.2 million apps, respectively [2]. If the app is predicted more accurately, users could get a more efficiently operational interface with the app to be used Another applicable case for app usage prediction is related to pre-loading the app into memory. The bandwidth has changeability depended on the time and locations of smartphone usage [2], [6] If they can predict the app usage, carriers could proactively determine the bandwidth allocation. We propose the WhatsNextApp approach, a deep learning approach based on Long Short-Term Memory Networks (LSTMs) that considers the sequences of app usage and is inspired by the concept of word embeddings. We predict the app usage only by using temporal features without including more Spatio-temporal features like location and smartphone activity as many state-of-the-art approaches do.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
