Toward A Human Behavior Models Anthology for Synthetic Agent Development

摘要

This paper describes an effort to foster the availability of Human Behavior Models / Performance Moderator Functions (HBM/PMFs) that the modeling and simulation community can use to increase the realism of their human behavior models. HBM/PMFs quantify the impact of human performance to internal and external stressors, and help to capture the role of personality and individual differences. To facilitate that process, we are creating a web-based anthology of HBM/PMFs that abstracts many 100s of them from diverse literatures, maps them into a taxonomy and common mathematical framework suitable for implementation, and assesses their validity and reuse issues. This paper reports on progress to date, anthology construction issues, and lessons learned.