Context Acquisition and Aggregation Supports the Realisation of Proactive Interaction: A Comparison between Decision Trees and Markov Models

Abstract

Over the last years, personal portable computer systems such as PDAs or laptops are being used in different contexts: at the office, during a meeting or at home. Switching between those contexts often requires the user to manually change differnt system settings, such as the audio volume, screen resolution, etc. In this work, we propose a first step towards a proactive user interface, predicting the next probable system changes, based upon previously learned sample. In particular, we will focus on two algorithms, decision trees ad Markov models, that may support this proactive interaction. By elaborating on some practical scenarios, we compare both implementations by evaluating their respective outcomes.