[2603.21653] MISApp: Multi-Hop Intent-Aware Session Graph Learning for Next App Prediction

Computer Science > Machine Learning arXiv:2603.21653 (cs) [Submitted on 23 Mar 2026] Title:MISApp: Multi-Hop Intent-Aware Session Graph Learning for Next App Prediction Authors:Yunchi Yang, Longlong Li, Jianliang Wu, Cunquan Qu View a PDF of the paper titled MISApp: Multi-Hop Intent-Aware Session Graph Learning for Next App Prediction, by Yunchi Yang and 3 other authors View PDF HTML (experimental) Abstract:Predicting the next mobile app a user will launch is essential for proactive mobile services. Yet accurate prediction remains challenging in real-world settings, where user intent can shift rapidly within short sessions and user-specific historical profiles are often sparse or unavailable, especially under cold-start conditions. Existing approaches mainly model app usage as sequential behavior or local session transitions, limiting their ability to capture higher-order structural dependencies and evolving session intent. To address this issue, we propose MISApp, a profile-free framework for next app prediction based on multi-hop session graph learning. MISApp constructs multi-hop session graphs to capture transition dependencies at different structural ranges, learns session representations through lightweight graph propagation, incorporates temporal and spatial context to characterize session conditions, and captures intent evolution from recent interactions. Experiments on two real-world app usage datasets show that MISApp consistently outperforms competitive baseline...