Autonomous User Profiling

  • Print

Problem Definition

 

For a human to trust an autonomous machine, the machine needs to understand the human. Similar to people, one person listens, look at the face of, and feels the other person. Understanding these cues are critical in the formation of trust and improving the interaction in a human social system. However, to profile a user autonomously is a challenging task.

 

Research Objectives

 

  • Designing autonomous computer agents that continuously assess human states.
  • Designing efficient data cleaning methods and cognitive features to understand human cues.
  • Identifying a robust set of features from EEG and physiological signals to understand human states and using these features to calibrate other methods for user profiling.