Nicholas Mark Gotts

Agent-based land-use models: a review of applications

Agent-based modelling is an approach that has been receiving attention by the land use modelling community in recent years, mainly because it offers a way of incorporating the influence of human decision-making on land use in a mechanistic, formal, and spatially explicit way, taking into account social interaction, adaptation, and decision-making at different levels. Specific advantages of agent-based models include their ability to model individual decision-making entities and their interactions, to incorporate social processes and non-monetary influences on decision-making, and to dynamically link social and environmental processes. A number of such models are now beginning to appear—it is timely, therefore, to review the uses to which agent-based land use models have been put so far, and to discuss some of the relevant lessons learnt, also drawing on those from other areas of simulation modelling, in relation to future applications. In this paper, we review applications of agent-based land use models under the headings of (a) policy analysis and planning, (b) participatory modelling, (c) explaining spatial patterns of land use or settlement, (d) testing social science concepts and (e) explaining land use functions. The greatest use of such models so far has been by the research community as tools for organising knowledge from empirical studies, and for exploring theoretical aspects of particular systems. However, there is a need to demonstrate that such models are able to solve problems in the real world better than traditional modelling approaches. It is concluded that in terms of decision support, agent-based land-use models are probably more useful as research tools to develop an underlying knowledge base which can then be developed together with end-users into simple rules-of-thumb, rather than as operational decision support tools.

Comparison of empirical methods for building agent-based models in land use science

The use of agent-based models (ABMs) for investigating land-use science questions has been increasing dramatically over the last decade. Modelers have moved from ‘proofs of existence’ toy models to case-specific, multi-scaled, multi-actor, and data-intensive models of land-use and land-cover change. An international workshop, titled ‘Multi-Agent Modeling and Collaborative Planning—Method2Method Workshop’, was held in Bonn in 2005 in order to bring together researchers using different data collection approaches to informing agent-based models. Participants identified a typology of five approaches to empirically inform ABMs for land use science: sample surveys, participant observation, field and laboratory experiments, companion modeling, and GIS and remotely sensed data. This paper reviews these five approaches to informing ABMs, provides a corresponding case study describing the model usage of these approaches, the types of data each approach produces, the types of questions those data can answer, and an evaluation of the strengths and weaknesses of those data for use in an ABM.

Representing and Reasoning with Qualitative Spatial Relations About Regions

Qualitative Reasoning (QR) has now become a mature subfield of AI as its tenth annual international workshop, several books (e.g. (Weld and de Kleer, 1990; Faltings and Struss, 1992)) and a wealth of conference and journal publications testify. QR tries to make explicit our everyday commonsense knowledge about the physical world and also the underlying abstractions used by scientists and engineers when they create models. Given this kind of knowledge and appropriate reasoning methods, a computer could make predictions and diagnoses and explain the behavior of physical systems in a qualitative manner, even when a precise quantitative description is not available or is computationally intractable. Note that a representation is not normally deemed to be qualitative by the QR community simply because it is symbolic and utilizes discrete quantity spaces but because the distinctions made in these discretizations are relevant to high-level descriptions of the system or behavior being modeled.

Constraint Networks of Topological Relations and Convexity

This paper studies the expressivity and computational complexity of networks of constraints of topological relations together with convexity. We consider constraint networks whose nodes are regular regions (a regular region is one equal to the closure of its interior) and whose constraints have the following forms: (i) the eight “base relations” of [12], which describe binary topological relations of containment and adjacency between regions; (ii) the predicate, “X is convex.” We establish tight bounds on the computational complexity of this language: Determining whether such a constraint network is consistent is decidable, but essentially as hard as determining whether a set of comparable size of algebraic constraints over the real numbers is consistent. We also show an important expressivity result for this language: If r and s are bounded, regular regions that are not related by an affine transformation, then they can be distinguished by a constraint network. That is, there is a constraint network and a particular node in that network such that there is a solution where the node is equal to r, but no solution where the node is equal to s.

Transient and asymptotic dynamics of reinforcement learning in games

Abstract Reinforcement learners tend to repeat actions that led to satisfactory outcomes in the past, and avoid choices that resulted in unsatisfactory experiences. This behavior is one of the most widespread adaptation mechanisms in nature. In this paper we fully characterize the dynamics of one of the best known stochastic models of reinforcement learning [Bush, R., Mosteller, F., 1955. Stochastic Models of Learning. Wiley & Sons, New York] for 2-player 2-strategy games. We also provide some extensions for more general games and for a wider class of learning algorithms. Specifically, it is shown that the transient dynamics of Bush and Mostellers model can be substantially different from its asymptotic behavior. It is also demonstrated that in general—and in sharp contrast to other reinforcement learning models in the literature—the asymptotic dynamics of Bush and Mostellers model cannot be approximated using the continuous time limit version of its expected motion.

What every agent-based modeller should know about floating point arithmetic

Floating point arithmetic is a subject all too often ignored, yet, for agent-based models in particular, it has the potential to create misleading results, and even to influence emergent outcomes of the model. Using a simple demonstration model, this paper illustrates the problems that accumulated floating point errors can cause, and compares a number of techniques that might be used to address them. We show that inexact representation of parameter values, imprecision in calculation results, and differing implementations of mathematical expressions can significantly influence the behaviour of the model, and create issues for replicating results, though they do not necessarily do so. None of the techniques offer a failsafe approach that can be applied in any situation, though interval arithmetic is the most promising.

Ontologies for transparent integrated human-natural system modelling

We propose an approach to modular agent-based land use modelling, based on ontologies in their computer science sense: formal representations of conceptualisations. The approach is primarily aimed at addressing the issue of model transparency. Human-natural systems models involve large numbers of submodels, making them difficult to understand for those not involved in their construction. We show that using ontologies to represent the structure and state of a simulation model improves transparency in two ways: First, the information about the structure and state is decoupled from the simulation software and can be independently processed. Second, the logics on which ontologies are based reflect more commonsense understandings of the relationships among concepts than those of computer programming languages.

e-Social Science and Evidence-Based Policy Assessment

The PolicyGrid project is exploring the role of Grid, Semantic Web, and Web 2.0 technologies to support e-Social Science, with particular emphasis on tools to facilitate evidence-based policy making. In this article, we discuss the challenges associated with construction of a provenance framework to support evidence-based policy assessment. We then discuss ourSpaces, a virtual research environment for e-Social Science that uses the Web 2.0 paradigm as well as Semantic Grid technologies and which provides researchers with facilities for management of digital resources using a novel natural language interface.

Enhancing workflow with a semantic description of scientific intent

In the e-Science context, workflow technologies provide a problem-solving environment for researchers by facilitating the creation and execution of experiments from a pool of available services. In this paper we will show how Semantic Web technologies can be used to overcome a limitation of current workflow languages by capturing experimental constraints and goals, which we term scientists intent. We propose an ontology driven framework for capturing such intent based on workflow metadata combined with SWRL rules. Through the use of an example we will present the key benefits of the proposed framework in terms of enriching workflow output, assisting workflow execution and provenance support. We conclude with a discussion of the issues arising from application of this approach to the domain of social simulation.

Semantic support for computational land-use modelling

In this paper we explore the use of proposed semantic grid standards and methodology through deployment of a land use modelling service. The FEARLUS-G service architecture is presented which allows large scale simulation experiments to be distributed over the grid. We also discuss ontology support for simulation parameters, hypotheses and results that facilitates sharing and re-use of such resources among land-use scientists. This leads to a description of infrastructure for semantic data management which integrates Jena2 and the ELDAS data access service.