Modeling erroneous human behavior: A context-driven approach.
Published: 08/2018
Wilson C
Abstract:
Abstract:
Artificial agents that model aspects of human behavior often model behaviors that an observer would regard as normal. In recent years, agents that exhibit observable erroneous behaviors have become common in a variety of applications, including simulated impaired agents to test assistive technologies and realistic agents in video games. In this paper, we present a context-driven approach to modeling plausible human behavior and a framework for modeling erroneous behavior which focuses on impairing an agent’s ability to recognize and deal effectively with anticipated contextual changes.
Citation: Wilson, C., & Turner, R.M. (2015). Modeling erroneous human behavior: A context-driven approach. In Modeling and Using Context: Proceedings of the Ninth International and Interdisciplinary Conference (Chistiansen, H., Stojanovic, I., & Papadopoulos, G.A., eds.) Volume 9405 of Lecture Notes in Artificial Intelligence, pp. 538-543, Springer.
URL: MaineSAIL.umcs.maine.edu/publications/papers/2015/wilson-turner15.pdf.