Discovering Agent Models using Process Mining: Initial Approach and a Case Study

Abstract

Agent-based modeling is widely used for modeling and simulation of self-organizing sociotechnical systems that are composed of distributed autonomous agents. In these systems, macro level behaviors emerge from local micro level behaviors of agents that follow rules and interact with each other and the environment. Although the individual agents' behaviors are typically described by sets of simple rules, the many interactions, heterogeneous populations, and complex topologies can make it challenging, or even impossible, to predict or steer the emergent behaviors beyond micro levels. Hence, the actual behaviors of such systems are generally hard to know beforehand, and they need to be observed to extract realistic models. In this paper, we propose a proof-ofconcept approach to discover agents' underlying models from log data generated from their behaviors, utilizing process mining. To conceptualize and demonstrate our initial approach, we use an illustrative example of the popular Schelling's model of segregation. Our findings provide encouraging initial evidence on how agent models can be extracted utilizing process mining techniques.

Using Process Mining for Face Validity Assessment in Agent-Based Simulation Models: An Exploratory Case Study

“…Figure 1 outlines the research design, which consists of several phases. This illustration is adapted from the approach proposed by [6]. In the following part, we provide a detailed account of the execution of each phase.…”

“…They show a form of agents' self-learning capabilities by incorporating knowledge obtained from process models into the agent decisions. In [6], the authors propose an approach to extract agents' underlying models from log data generated from their behaviors, utilizing process mining. The authors demonstrate this approach using the Schelling model of segregation, showing how agent models can be extracted utilizing process mining techniques.…”

“…The literature also reports examples of model extensions (see e.g., [22,32,34]). Additionally, this study aims to build upon the research conducted by [6] by using their case study as a foundation for further investigation.…”

“…Model implementation. For the implementation of our simulation model, we based our design on the approach proposed by [6]. Additionally, we have used an existing simulation implementation developed by [13] for verification purposes.…”

“…However, available studies do not cover agent-based simulations that are characterized by for example numerous interactions, heterogeneous populations, and complex topologies. Additionally, the evolving nature of agent behavior and their available knowledge are often overlooked [4,6], raising concerns about model validity. Leveraging process mining techniques to support face validity assessments can enhance the validity of agent-based simulation models.…”