Real-Time Passive Data for Study of Individual and Interpersonal Behaviors

Project Team

Byron Reeves, PhD
C. Lee Giles, PhD

The availability of mobile real-time data streams (e.g., laptops, smart phones, etc.) make this an exciting time. These new data are allowing us to re-conceptualize how we live, learn, communicate, and organize our daily lives to achieve personal goals. This project is about the interdependence of behavior across multiple domains (school, health, leisure), multiple time-scales (days, months), and multiple spatial/virtual locations (classroom, gym, home, games, media). We focus on the individual – the quantified self – as a complex system that exchanges data with many different data archives. Changes in how individuals multi-task means that understanding individuals’ behavior will depend as much on stitching together experiences scattered across multiple domains as on examining behavior within any single domain (e.g., school, health, social relationships). Data analysis plans integrate “top-down” (deductive theory testing) and “bottom-up” (inductive data mining) methods developed specifically for analysis of individual-level multivariate data.  We will identify the features, sequences, and time-scales that facilitate prediction of specific types of intra-personal and inter-personal dynamics.