For a methodology to be useful to science, it should provide two things: first a means of explanation, and second, a mechanism for improving that explanation. Agent Based Modelling (ABM) is a method that facilitates exploring the collective effects of individual action selection. The explanatory force of the model is the extent to which an observed meta-level phenomena can be accounted for by the behaviour of its micro-level actors. This article demonstrates this methodology can be applied to the biological sciences, that agent-based models like any scientific hypotheses can be tested, critiqued, generalised or specified. We review the state of the art for ABM as a methodology for biology. We then present a case study based on the most widely-published agent-based model in the biological sciences: Hemelrijk’s DomWorld, a model of primate social behaviour (Hemelrijk, 1999b, 2002b, 2004). Our analysis shows some significant discrepancies between the model and the behaviour of the species we compare it to, macaques. We alsodemonstrate that the model is not fragile: its other results are still valid and can be extended to compensate for these problems. This is a standard advantage of experiment-based AI modelling techniques over analytic modelling.

The models used in social simulation to date have mostly been very simplistic cognitively, with little attention paid to the details of individual cognition. This work proposes a more cognitively realistic approach to social simulation. It begins with a model created by Gilbert (1997) for capturing the growth of academic science. Gilbert’s model, which was equation-based, is replaced here by an agent-based model, with the cognitive architecture CLARION providing greater cognitive realism. Using this cognitive agent model, results comparable to previous simulations and to human data are obtained. It is found that while different cognitive settings may affect the aggregate number of scientific articles produced, they do not generally lead to different distributions of number of articles per author. The paper concludes with a discussion of the correspondence between our model and the constructivist view of academic science. It is argued that using more cognitively realistic models in simulations may lead to novel insights.

Abstract. This paper is devoted to exploring the relationships between computational agents, as they can be found in multi-agent systems (MAS) or Distributed Artificial Intelligence (DAI), and the different techniques regrouped under the generic appellation “multi-agent based simulation ” (MABS). Its main purpose is to show that MABS, despite its name, is in fact rarely based on computational agents. We base our demonstration on an innovative presentation of the methodological process used in the development of current MABS systems. This presentation relies on the definition of the different roles involved in the design process, and we are able to show that the notion of “agent”, although shared at a conceptual level by the different participants, does not imply a systematic use of computational agents in the systems deployed. We then conclude by discussing what the use of computational agents, based on the most interesting research trends in DAI or MAS, might provide MABS with.

The aim of this paper is to assist researchers in understanding the dynamics of simulation models that have been implemented and can be run in a computer, i.e. computer models. To do that, we start by explaining (a) that computer models are just input-output functions, (b) that every computer model can be re-implemented in many different formalisms (in particular in most programming languages), leading to alternative representations of the same input-output relation, and (c) that many computer models in the social simulation literature can be usefully represented as time-homogeneous Markov chains. Then we argue that analysing a computer model as a Markov chain can make apparent many features of the model that were not so evident before conducting such analysis. To prove this point, we present the main concepts needed to conduct a formal analysis of any time-homogeneous Markov chain, and we illustrate the usefulness of these concepts by analysing 10 well-known models in the social simulation literature as Markov chains. These models are: * Schelling's (1971) model of spatial segregation * Epstein and Axtell's (1996) Sugarscape * Miller and Page's (2004) standing ovation model * Arthur's (1989) model of competing technologies * Axelrod's (1986) metanorms models * Takahashi's (2000) model of generalized exchange * Axelrod's (1997) model of dissemination of culture * Kinnaird's (1946) truels * Axelrod and Bennett's (1993) model of competing bimodal coalitions * Joyce et al.'s (2006) model of conditional association In particular, we explain how to characterise the transient and the asymptotic dynamics of these computer models and, where appropriate, how to assess the stochastic stability of their absorbing states. In all cases, the analysis conducted using the theory of Markov chains has yielded useful insights about the dynamics of the computer model under study.

Abstract. Multi-Agent Systems (MASs) is a promising approach for studying human societies. The field, however, still encounters major methodological difficulties. Due to the high level of complexity that can be attained using this type of model, an efficient way for selecting an appropriate level of detail for each specific context has become a necessity. The aim of this research is to present a new approach for automatic model abstraction. The developed technique is applied to Rule-Based (RB) MASs, one of the most common types of model in social simulation. The approach has been tested on well-known model examples and proved to give advantage by replacing detailed models with simpler ones that are suitable for specific contexts. 1

Abstract: Object-oriented (OO) programming has limitations when used to implement abstract multi-scale, spatially-integrated, agent-based models, that could potentially be addressed using relational databases (RDB). Although this would involve rethinking the approach to designing such models, the combined OO-RDB approach has a number of appealing advantages for multi-scale simulations, such as allowing the user rather than the programmer to specify the scale at which various land-use processes take place. It also provides a basis for a more realistic representation of the relationship between agents and their environment.

The objectives of this paper are to define and classify different types of errors and artefacts that can appear in the process of developing an agent-based model, and to propose activities aimed at avoiding them during the model construction and testing phases. To do this in a structured way, we review the main concepts of the process of developing such a model – establishing a general framework that summarises the process of designing, implementing, and using agent-based models. Within this framework we identify the various stages where different types of errors and artefacts may appear. Finally we propose activities that could be used to detect (and hence eliminate) each type of error or artefact.

This document introduces AgeS - an "Agent System". The material contained in this introduction supplies some background that the reader may find helpful in later chapters. In particular, it gives a brief overview of distributed computing, client/server applications, naming services, agents, and agent systems

doi:10.1098/rstb.2007.2061

Abstract. Now that complex Agent-Based Models and computer simulations spread over economics and social sciences- as in most sciences of complex systems-, epistemological puzzles (re)emerge. We introduce new epistemological tools so as to show to what precise extent each author is right when he focuses on some empirical, instrumental or conceptual significance of his model or simulation. By distinguishing between models and simulations, between types of models, between types of computer simulations and between types of empiricity, section 2 gives conceptual tools to explain the rationale of the diverse epistemological positions presented in section 1. Finally, we claim that a careful attention to the real multiplicity of denotational powers of symbols at stake and then to the implicit routes of references operated by models and computer simulations is necessary to determine, in each case, the proper epistemic status and credibility of a given model and/or simulation. Keywords: Agent-Based models, simulation, social sciences, economics, epistemology, conceptual exploration, experiment, iconicity, denotational