Abstract

The concept of ubiquitous computing, as introduced by Marc Weiser in the early 90s, spurred research into various kinds of context-aware systems and applications. There is a wide range of contextual parameters, including location, time, temperature, devices and people in proximity, which have been part of the initial ideas about context-aware computing. While locational context is already a well understood concept, social context—-based on the people around us—-proves to be harder to grasp and to operationalize. In this talk, I will explore the concept of social context in ubiquitous computing. The focus lies on the measurement of social context in urban environments with the Bluetooth device discovery protocol, and a number of methods to visualize personal behavior, group behavior, and the behavior of groups in relation to places. These methods are based on the concept of structural equivalence in social network analysis. I will present a number of visualizations to understand social context in relation to location and time—-both on a long term and short term. My research builds on a new dataset collected for this purpose, which contains Bluetooth contact data from an egocentric perspective in a number of different social settings, some of which are complemented by data from stationary Bluetooth scanners and Cell-ID measurements.