Enabling pervasive applications by understanding individual and community behaviors. (Nouvelles applications pervasives par la modélisation des comportements individuels et communautaires)

With the rapid advancement of sensing, computing, networking and storage, recent years have been witnessing a phenomenal growth of sensor-embedded mobile phones and prevailing use of GPS devices in vehicles. The digital footprints collected from such systems provide novel ways to perceive an individual’s behaviors. Furthermore, large collections of digital footprints from communities bring novel understandings of human behaviors from the community perspective (community behaviors), such as investigating their characteristics and learning the hidden human intelligence. The perception of human behaviors from the sensing digital footprints enables novel applications for the sensing systems. However, it’s not easy to recognize individual and community behaviors from the digital footprints. Firstly, there are various problems in the sensing data that need to be preprocessed. Secondly, there are big gaps to be bridged between the raw sensing data and the recognition of human behaviors. Thirdly, with the observation of behaviors in a community, learning the hidden human intelligence is a complicated problem. Lastly, for a smart sensing system that monitors the behaviors of a large number of individuals, it’s often hard to be scalable and support real time responses. Bases on the digital footprints collected with accelerometer-embedded mobile phones and GPS equipped taxis, in this dissertation we present our work in recognizing individual behaviors, capturing community behaviors and demonstrating the novel services enabled. In recognizing individual behaviors, we present the recognition of an individual’s physical activities with the accelerometer readings collected from mobile phones placed in the pockets around the pelvic region. Particularly, to overcome the variance of location and orientation problem, we introduce the orientation invariant sensing magnitude to the sensor readings and improve the recognition accuracy about 8 percent. Besides, we also reduce the computing cost with a feature reduction method. With the GPS footprints of a taxi, we summarize the individual anomalous passenger delivery behaviors and improve the recognition efficiency of the existing method iBOAT by introducing an inverted index mechanism. Besides, based on the observations in real life, we propose a method to detect the work-shifting events of an individual taxi. With real-life large-scale GPS traces of thousands of taxis, we investigate the anomalous passenger delivery behaviors and work shifting behaviors from the community perspective and exploit taxi serving strategies. We find that most anomaly behaviors are intentional detours and high detour inclination won’t make taxis the top players. And the spatialtemporal distribution of work shifting events in the taxi community reveals their influences. While exploiting taxi serving strategies, we propose a novel method to find the initial intentions in passenger finding. With the strategies modeled for thousands of drivers, we isolate the good and bad ones by learning from good drivers, measuring the correlation between each strategy and performance, and classifying different driver groups. Furthermore, we present a smart taxi system as an example to demonstrate the novel applications that are enabled by the perceived individual and community behaviors. te l-0 08 14 60 4, v er si on 1 17 A pr 2 01 3