An integrated approach for the validation of emergence in component-based simulation models

Abstract

Emergent properties are becoming increasingly important as systems grow in size and complexity. Despite recent research interest in understanding emergent behavior, practical approaches remain a key challenge. This paper proposes an integrated approach for the identification of emergence with two perspectives. A post-mortem emergence analysis requires a-priori knowledge about emergence and can identify the causes of emergent behavior. In contrast, a live analysis, in which emergence is identified as it happens, does not require prior knowledge and relies on a more rigorous definition of individual model components in terms of what they achieve, rather than how. Our proposed approach integrates reconstructability analysis in the validation of emergence included in our proposed component-based model development life-cycle.

Formalization of emergence in multi-agent systems

“…Verifying emergence without prior knowledge of emergence is a research challenge. Validation allows us to determine whether an emergent property is beneficial or harmful [37]. Finally, reasoning enables the understanding of the cause-and-effect of emergence.…”

Post-mortem analysis of emergent behavior in complex simulation models

“…Despite ongoing research since the 1970s [1,7,11,15,26], very few methods for identifying, classifying, and explaining emergent properties exist [5,18,26,27], and they are usually employed only on simplified examples that are not often found in real life [27]. These approaches can be classified broadly from two orthogonal perspectives, namely live analysis and post-mortem analysis [27].…”

“…These approaches can be classified broadly from two orthogonal perspectives, namely live analysis and post-mortem analysis [27]. Live emergence analysis proposes to identify emergence as it happens, using meta-models of calculated composed model states [27], or representations of system interaction [4].…”

“…These approaches can be classified broadly from two orthogonal perspectives, namely live analysis and post-mortem analysis [27]. Live emergence analysis proposes to identify emergence as it happens, using meta-models of calculated composed model states [27], or representations of system interaction [4]. Post-mortem emergence analysis starts from a definition of a known or observed emergent property and aims to identify its cause, in terms of model components and their interaction [5,26].…”

Understanding complex systems: Using interaction as a measure of emergence

“…The above approaches are demonstrated using simple models such as flocks of birds or predator-prey but have limiting assumptions and constraints when applied to more complicated systems. For example, most approaches do not consider mobile agents (Kubik 2003), assume unfeasible a-priori specifications and definitions of emergent properties (Szabo and Teo 2012), or do not scale beyond models with a small number of agents (Chen et al 2007, Kubik 2003. In the multi-agent systems community, approaches focus more on the engineering of systems to exhibit beneficial emergent behavior and less on its identification and validation (Bernon et al 2003, Jacyno et al 2009, Salazar et al 2011).…”