Center for Human Behavioral Informatics
Organizers: Shri Narayanan (Coordinator), Richard Leahy, Krishna Nayak, Urbashi Mitra, Andy Mollisch, Antonio Ortega, Jay Kuo, Justin Haldar, Panayiotis Georgiou
Sponsored: Spring 2012 – present

The Center for Human Behavioral Informatics will create and use engineering advances to the study of human behavior. The scientific thrust will explore the gaps in our understanding of the complex interplay between the human brain, body and behavior facilitated by engineering advances in sensing, signal processing and systems modeling. The computational and technological thrusts would focus on creating novel techniques and tools for especially for supporting health diagnostics and interventions.

Human behavior is complex and multifaceted. It manifests an intricate interplay between the human mind, brain and the body. Importantly, not only does it represent the natural dynamics of an individual’s internal neuro, cognitive and physiological state, it reflects the influence of other agents and the environment. Behavioral expressions of an individual’s actions and or interactions hence can be widely varied depending on the individual’s state, the nature of the task engaged in, as well as the external influences and context. This heterogeneity can arise from, and be attributed to a wide range of factors ranging from an individual’s age, gender, socio-cultural background to their physical and mental health status and abilities, including possible differences due to illness, disease or disability. Additional variability in human behavior displays arise from the complex interplay between these and the infinite sources of variability in the cognitive demands of the tasks and activities, and variability in the contextual factors and the environment, including the behavior of other individuals. All these factors make the understanding and automatic decoding of human behavior cues is a vastly challenging engineering problem, from sensing to modeling.

The proposal to MHI is to organize a steering group to establish an engineering center at USC on Human Behavioral Informatics. The specific goals are to:
•define and refine the goals and scope of the center
•create a compelling and cohesive team drawing colleagues from within and outside USC
•identify one or two concrete major demonstration projects
•identify specific extramural funding opportunities

Understanding, describing, and influencing, human behavior is central to many domains of human endeavor. This could relate to understanding normative (“typical”) behavior patterns of an individual engaged in a task or activity. More often, it relates to detecting, analyzing and modeling behavior deviation from what is deemed typical, and in factoring the source attributable to this variability. Critically, no other domain exemplifies the centrality of behavioral analysis and modeling than that related to human mental health and well being. In particular, research and practice in psychology and psychiatry focuses on diagnosing, managing and treating atypical and distressed behavior by eliciting and or observing behavioral cues and patterns. Applications to numerous other domains where monitoring and tracking behavior in real world, so called “free living”, settings can transform how we understand and respond to the underlying problems include metabolic health monitoring, mild cognitive impairment and onset of dementia and Alzheimers, post stroke recovery, and other quality of life based applications targeting the elderly.