Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Different Facets for Different Experts: A Framework for Streamlining The Integration of Qualitative Insights into ABM Development

Vivek Nallur, Pedram Aghaei, Graham Finlay

TL;DR

The paper tackles the difficulty of integrating qualitative insights from diverse experts into ABMs, proposing a facet-based architecture that decouples programming from qualitative inputs. It introduces roles for ABM Programmers, Domain Experts, and End-Users, with a graph-based BehaviourFlow to encode context-sensitive decision-making and a web-enabled Policy/Scenario Design system for rapid experimentation. Key contributions include the facet framework, BehaviourFlow with BehaviorTriggers, and a web-based interface for policy and scenario management, demonstrated in an economic migration use case. This approach enables asynchronous, evolving qualitative inputs, supports comparisons across multiple expert perspectives, and maintains separation from core code, thereby enhancing stakeholder involvement and model fidelity in socially complex simulations.

Abstract

A key problem in agent-based simulation is that integrating qualitative insights from multiple discipline experts is extremely hard. In most simulations, agent capabilities and corresponding behaviour needs to be programmed into the agent. We report on the architecture of a tool that disconnects the programmed functions of the agent, from the acquisition of capability and displayed behaviour. This allows multiple different domain experts to represent qualitative insights, without the need for code to be changed. It also allows a continuous integration (or even change) of qualitative behaviour processes, as more insights are gained. The consequent behaviour observed in the model is both, more faithful to the expert's insight as well as able to be contrasted against other models representing other insights.

Different Facets for Different Experts: A Framework for Streamlining The Integration of Qualitative Insights into ABM Development

TL;DR

The paper tackles the difficulty of integrating qualitative insights from diverse experts into ABMs, proposing a facet-based architecture that decouples programming from qualitative inputs. It introduces roles for ABM Programmers, Domain Experts, and End-Users, with a graph-based BehaviourFlow to encode context-sensitive decision-making and a web-enabled Policy/Scenario Design system for rapid experimentation. Key contributions include the facet framework, BehaviourFlow with BehaviorTriggers, and a web-based interface for policy and scenario management, demonstrated in an economic migration use case. This approach enables asynchronous, evolving qualitative inputs, supports comparisons across multiple expert perspectives, and maintains separation from core code, thereby enhancing stakeholder involvement and model fidelity in socially complex simulations.

Abstract

A key problem in agent-based simulation is that integrating qualitative insights from multiple discipline experts is extremely hard. In most simulations, agent capabilities and corresponding behaviour needs to be programmed into the agent. We report on the architecture of a tool that disconnects the programmed functions of the agent, from the acquisition of capability and displayed behaviour. This allows multiple different domain experts to represent qualitative insights, without the need for code to be changed. It also allows a continuous integration (or even change) of qualitative behaviour processes, as more insights are gained. The consequent behaviour observed in the model is both, more faithful to the expert's insight as well as able to be contrasted against other models representing other insights.
Paper Structure (15 sections, 6 figures)

This paper contains 15 sections, 6 figures.

Table of Contents

  1. Introduction
  2. Integrating Qualitative Information Into ABMs
  3. Common Problems
  4. Typical Integration Approaches
  5. Separated Experts
  6. Architecture
  7. Facets (Role: ABM Programmer)
  8. BehaviourFlow (Role: Domain Expert)
  9. Policy (Role: Policy Maker / End-User)
  10. Scenario Design (Role: End-User)
  11. Domain Expert Influencing Agent Behaviour
  12. Understanding the BehaviourFlow
  13. User Roles and Activities
  14. Known Weaknesses
  15. Conclusion

Figures (6)

  • Figure 1: How new facets restructure the model
  • Figure 2: A sample BehaviourFlow for a MigrantAgent procuring documents, with the sequence showing the dependency between documents
  • Figure 3: Creating new policy by defining conditions and actions
  • Figure 4: Components of the Facet Framework
  • Figure 5: A trigger function to define the possibility or desire to apply for a new job
  • ...and 1 more figures