Automated Cognitive Behavior Analysis

Understanding and predicting human behavior has always been a fundamental research question in AI. This project proposes a methodology to address that problem termed automated cognitive behavior analysis (ACBA). The methodology allows uncovering the underlying behavior structures — the mental plans guiding and shaping people’s actions carried out in the world — given observed behavior performed in the context of complex tasks. It involves the use of genetic programming (GP) to iteratively generate programs capable of explaining the behavior exhibited by an individual in a given task. It also includes a set of tools to help analyze and interpret the invariant cognitive structures responsible for different observed behaviors.

[bibshow file=my-publications.bib show_links=1 format=custom-ieee template=custom-bibshow highlight=”P. Sequeira”]

The ACBA methodology uses GP as hypothesis generation and testing of human behavior. It generates programs by combining symbols encoding relevant task information and certain operators, given some behavior instance of interest. A program’s fitness is measured according to how consistent its output is in relation to the outcomes of the exhibited behavior. By using standard GP techniques, ACBA progressively generates candidates attaining higher and higher degrees of fitness. When GP finishes, the most fit programs are chosen as hypotheses of, i.e., possible solutions for, the underlying structure of the observed behavior.

The work in [bibcite key=sequeira2018cogsci] studies the influence of personality traits in negotiation by using ACBA. The goal is to study the behavior exhibited by individual during a multi-level, multi-issue, sequential bargaining task against an artificial agent. The study focuses on the influence of the personality traits of social-value orientation (SVO) and Machiavellianism (Mach). The results show that ACBA is able to identify differences in the outcomes of programs emerging from GP that are consistent with the influences that different SVO and Mach profiles have in human negotiation behavior, as informed by specialized literature in negotiation behavior.

Related Publications