Operation Prediction for Context-Aware User Interfaces of Mobile Phones

Nowadays mobile phones are multifunctional devices that provide us with various useful applications and services anytime and anywhere. However, people are sometimes unable to access an appropriate application due to the complexity and depth of the menu structure. This paper focuses on a feasibility study of operation prediction using observable attributes to realize self-optimization functionality in the mobile phones that can automatically and adaptively change their user interface (UI) according to user characteristics and circumstances. Machine learning (ML) is a promising technology for enhancing UI. However, few studies have been conducted for the operation prediction using the ML framework. We analyzed the real usage data collected by practical mobile phones and found that ML-based prediction methods were feasible to estimate future operations, and to provide context-aware UI.