Context-aware mobile computing: learning context- dependent personal preferences from a wearable sensor array

Context-aware computing describes the situation where a wearable/mobile computer is aware of its user's state and surroundings and modifies its behavior based on this information. We designed, implemented, and evaluated a wearable system which can learn context-dependent personal preferences by identifying individual user states and observing how the user interacts with the system in these states. This learning occurs online and does not require external supervision. The system relies on techniques from machine learning and statistical analysis. A case study integrates the approach in a context-aware mobile phone. The results indicate that the method is able to create a meaningful user context model while only requiring data from comfortable wearable sensor devices.

1 Introduction

Our goal is to make the mobile computer capable of sensing the users and their current state, exploiting context information to significantly reduce demands on human attention. To minimize user distraction, a pervasive computing system must be context-aware [1]. By estimating user context qualities such as location, activity, physiology, schedule, and ambient context information, a human's current state can be determined to a level which allows configuring the settings of a mobile device proactively. Wrong configuration can often lead to frustrating conse-quences, such as unwanted interruptions or additional work to undo the system's decisions. In addition, mobile computing applications face challenges such as varying resource availability. Any sensor data should be captured by a body-worn sensor array; wearable hardware must pro-vide enough processing capability for data analysis and context inference.