Social motion : mobile networking through sensing human behavior

Abstract

Low-level sensors can provide surprisingly high-level information about social interactions. The goal of this thesis is to define the components of a framework for sensing social context with mobile devices. We describe several sensing technologies - including infrared transceivers, radio frequency scanners, and accelerometers - that both capture social signals and meet the design constraints of mobile devices. Through the analysis of several large datasets, we identify features from these sensors that correlate well with the underlying social structure of interacting groups of people. We then detail the work that we have done creating infrastructure that integrates social sensors into social applications that run on mobile devices.