@MASTERSTHESIS\{IMM2013-06556,
    author       = "P. Sapiezynski",
    title        = "Measuring Patterns of Human Behavior through Fixed-Location Sensors",
    year         = "2013",
    school       = "Technical University of Denmark, {DTU} Compute, {E-}mail: compute@compute.dtu.dk",
    address      = "Matematiktorvet, Building 303{-B,} {DK-}2800 Kgs. Lyngby, Denmark",
    type         = "",
    note         = "{DTU} supervisors: Jakob Eg Larsen, jaeg@dtu.dk, and Sune Lehmann J{\o}rgensen, {DTU} Compute",
    url          = "http://www.compute.dtu.dk/English.aspx",
    abstract     = "Fixed-location sensor networks are growing in popularity as they provide a costeffective solution for measuring behavior of large populations. In many cases, existing infrastructure can be turned into a sensor network by using custom software; alternatively, off-the-shelf hardware can be used to build such networks from scratch. Since collecting data through such sensors does not require any interaction with the observation subjects, it does not influence their behavior, and can thus provide objective insights. This very same characteristic makes fixed-location sensing controversial in terms of privacy protection, as it can be deployed without any consent from the observed persons.
This thesis provides a general review of the fixed-location sensor networks in use currently, and the state of the art statistical approaches to modeling the collected data. It discusses possible directions of future development of fixed-location sensing as well as privacy and security implications of such approach
to behavioral tracking. Finally, analysis of real world data collected during experimental deployments in different contexts and the fixed-location sensor perspective is compared to mobile sensor perspective, to show the following:
- It is possible to capture mobility and interaction traces of crowd participants through stationary Bluetooth scanners; modeling such data using Infinite Relational Model reveals underlying motivations of the participants.  
- Stationary WiFi Access Points provide a highly accurate insight into social networks based on physical proximity (who spends time with whom); it is possible to extract behavioral patterns from such data and, through  mathematical modeling, recover the cognitive social network (who likes whom).
Additionally, the thesis presents one paper, which I authored with researchers from Alan Mislove’s Theory and Networked Systems Lab and David Lazer’s LazerLab at Northeastern University, which however falls outside of the main scope of the thesis."
}