Introduction to the special issue on “Human Behavior in Ubiquitous Environments: Modeling of Human Mobility Patterns” (original) (raw)

2010, Pervasive and Mobile Computing

Introduction to the special issue on ''Human Behavior in Ubiquitous Environments: Modeling of Human Mobility Patterns'' Ubiquitous computing requires seamless access to media, information sharing and communication through heterogeneous systems, which are often distributed for example deployed on mobile devices or deeply embedded in the physical environment. Recent advances in computing technology allow researchers and practitioners to realize at least in part this vision of ubiquitous computing and build large-scale systems thanks to the widespread availability of more and more powerful mobile computing platforms. These devices have become an essential part of the everyday experience for billions of people. For this reason, a key aspect is the design and implementation of systems built around the life of individuals. Thus, the understanding of human behavior-and, in particular, its impact on technological issues relevant to these systems-has emerged as a fundamental research area in ubiquitous computing. In this first of two special issues of Pervasive and Mobile Computing on human behavior in ubiquitous environments we present state-of-the-art research contributions in this fundamental and exciting field. In particular, the work presented in this issue relates to models of human behavior with special focus on human mobility and the inference of particular activities. Smart-phones are emerging as a de-facto standard computing platform for ubiquitous context-aware systems, since they are able to continuously sense the dynamic context of users. In ''Predicting mobility events on personal devices'' Peddemors et al. propose a method for the prediction in time of the next occurrence of a context event of interest. More specifically, the authors focus on the prediction of network visibility events as observed through the wireless network interfaces of mobile devices. The approach is based on a predictor that analyses the event stream for forecasting context changes. Using two real world data sets, the authors found that including predictors of infrequently occurring events results in better predictions. They also prove that cross-sensor and cross-interface information in most cases improves the prediction performance. Mobility is also the key theme of the following paper, ''The co-evolution of taxi drivers and their in-car navigation systems''. In this article Girardin and Blat study how the adoption of in-car navigation systems changed the practice of the community of taxi drivers of Barcelona, Spain, from an ethnographical perspective. The results show co-evolution: taxi drivers adapt to their in-car navigation systems and, at the same time, they adapt them to their needs. In particular, the authors found evidence of an alteration of the learning processes to reduce stress rather than to improve efficiency. These results can be applied to the design of the next generation of this class of systems. Location information plays a key role in many mobile applications, from context-based search and advertising to sensing systems. In ''Mobility Profiler: A Framework for Discovering Mobility Profiles of Cell Phone Users'', Ali Bayr et al. present Mobility Profiler, a framework for the discovery of mobile cellular phone user profiles starting from cell based location data. To validate their results, the authors use real-world cell phone log data and report results about frequent mobility patterns and profiles, showing that user residence time follows a long tail distribution. The following paper by Kaltenbrunner et al. is also about the analysis of human movement, again from measurements in Barcelona, Spain. In ''Urban cycles and mobility patterns-Exploring and predicting trends in a bicycle-based public transport system'' the authors provide study human mobility in an urban area by analyzing the amount of available bikes in the stations of the community bicycle program ''Bicing'' in Barcelona. In fact, by means of data sampled from the operator's website, they are able to detect temporal and geographic mobility patterns within the city. These patterns are applied to predict the number of available bikes for any station in advance. In ''Utilization of User Feedback in Indoor Positioning System'' Hossain et al. propose an interpolation-based fingerprinting technique exploiting user feedback which does not require the typical exhaustive training phase of the existing indoor localization solutions. The authors show that users' feedback can be used for fine-tuning an under-trained positioning system with filtering. The experimental results also demonstrate that the participation of end-users can actually assist in the incremental construction of a positioning system. The authors show that the proposed user feedback-based positioning system can dynamically adapt to context changes.