文摘
Inferring social relationships among individuals is a difficult problem, typically requiring either intrusive measurements (e.g., surveys) or convenience samples (e.g., ties captured by online social networking systems). Automatically collected activity data from mobile devices or similar sources provide a promising alternative for inexpensive and unobtrusive detection of interpersonal relationships. Here, we introduce a new method—activity correlation spectroscopy—for inferring relationships by exploiting the spectral and distributional structure of activity correlation within dyads. Unlike existing techniques, our approach can be employed with minimal, individual-level (i.e., non-relational), and non-identifying data that are easily collected using commodity hardware. We demonstrate our methodology via an application to detection of friendship and group co-membership using mobile device and survey data from the MIT Reality Mining study (Eagle et al. in Proc Natl Acad Sci 106:15274–15278, 2009).