EURO 2024 Copenhagen
Abstract Submission

EURO-Online login

1304. How, when and why human agents act in a simulation? The PRIAGE architecture.

Invited abstract in session WA-7: Behaviour and decision processes , stream Behavioural OR.

Wednesday, 8:30-10:00
Room: 1019 (building: 202)

Authors (first author is the speaker)

1. Miguel Baigorri
Statistics, informatics and Mathematics, Public University of Navarra
2. Fermin Mallor
Institute of Smart Cities, Public University of Navarre

Abstract

Modeling human behavior is a highly complex task because humans don't always act rationally, and their actions can be influenced by a wide variety of stimuli that are difficult to measure. However, accurately modeling human behavior is one of the main challenges in developing useful agent-based simulations that can be applied to real-life situations.
We propose a new architecture called PRIAGE (Pursued, Requested, and Induced Actions for reaching Goals in the Environment), which offers an innovative approach to the influence of an agent's internal state on their actions. In the majority of existing human-behavior paradigms the agent’s internal state is used only to determine which action to choose from a set of possible actions. PRIAGE focuses on describing the different reasons that can trigger an action: agents’ actions are either requested by the environment, induced by the agent's internal state, or actively executed by the agent to achieve goals. A second novelty is that the internal state is not only used to determine what action to do next, but may also influence how well the action is performed, creating performance measures that can modify the environment and the agent.
After describing the new proposed architecture, we apply it to a case study analyzing physicians in an emergency department. We focus on how the internal state of the agent influences two performance measures: the treatment time and treatment quality of patients.

Keywords

Status: accepted

Back to the list of papers