*CSIRO Division of Building, Construction and Engineering, Australia

Keywords: Agent-Based Modeling, Artificial Intelligence

Some pioneering attempts at agent-based modeling in the social sciences were carried out a generation ago by Harvard’s Thomas Schelling. In many respects, Schelling anticipated various themes that are currently at the forefront of agent-based simulation, social complexity, and economic evolution. Yet he lacked the computational capacity to reveal the full consequences of his innovative ideas. Because the study of large numbers of agents, with changing patterns of interactions, quickly gets too difficult for an explicit mathematical solution, computer simulation becomes obligatory. Today’s agent-based experiments in silico have three ingredients: boundedly rational agents, an environment in which they "live", and a set of rules. Although these artificial agents are intelligent and adaptive, the processes of inference, learning and discovery that they can embrace are often simplistic. Learning agents can be arbitrarily intelligent, but unless they know other agents’ learning methods they cannot know if their own learning processes are efficient. Agents can only discover the efficacy of their own learning methods by testing them against others. In this paper, we review some examples of locally interacting agents who reason inductively to produce novel, large-scale effects. We cannot "solve" this kind of agent-based model, but instead we must "evolve" it. Individual beliefs become endogenous, competing within a larger ecology of all agents’ beliefs. Furthermore, this ecology of beliefs co-evolves over time, as agents adapt to the choices of other agents and the state of their environment. The paper concludes with a few suggestions concerning the ways we might build more realistic co-evolutionary learning mechanisms into agent-based experiments with artificial societies.

School of management UMIST, P.O.Box 88 Manchester M60 1QD, United Kingdom

Tel: 016 200 3430 Fax: 0161 200 3505, Email: j.freeman@umist.ac.uk

Keywords: Complex Societal Problems, Charitable Funding

Trends away from social service have left many communities increasingly reliant on charitable bodies for practical and financial assistance. As a consequence, local organizations - such as the Buxton & District Lions Club (BDLC) - have seen a marked hike in requests for help even though, in the case of BDLC, its funding capability in recent years has hardly changed.

Compounding the problem, applications historically have been so diverse that prioritizing which to act on has often been both difficult and extremely time-consuming.

As part of an investigation into how its selection process could be made more efficient (faster and scientific) as well as transparent, BDLC recently tested the AHP methodology out on a sample of past applications. Despite initial reservations, the technique in fact proved remarkably insightful and predictions from the analysis were found to significantly agree with funding decisions actually taken.

Natural Resources Canada, 580 Booth Street, K1A 0E4, Ottawa Ontario, Canada

Tel: 613-947-9023, Fax: 613-947-909, Email: alsimard@nrcan.gc.ca

Keywords: Knowledge Management, Policy Analysis, Decision-Support Systems

This paper provides a comprehensive framework that incorporates all aspects of knowledge management in a policy analysis context. Social issues are substantially more complex and far less quantitative than those, which conventional Decision-Support Systems can address. Issue analysis requires processing information from multiple domains. Therefore, an infrastructure is needed to horizontally integrate data warehouses, information repositories, and knowledge sources across multiple domains. We must develop ways to use the revolution in information and computer technologies to access inputs; add value to content by transforming data into information and information into knowledge; and disseminate knowledge products. The technology must also be adapted to preserving knowledge assets. Issue analysis is also fundamentally an institutional undertaking. Therefore, knowledge management must address the organisational context that surrounds it. This includes: understanding knowledge work and knowledge workers; forms and structures of information governance; strategic planning and policy formulation in relation to knowledge; establishing and allocating resources to knowledge management initiatives; changing organisational cultures from controlling to sharing knowledge; and fostering organisational learning to increase, capture, and mobilise tacit knowledge. Finally, knowledge management must be promoted as a key component of policy analysis.

Center for Tele-Information, Technical University of Denmark

DK-2800 Lyngby, Denmark, Tel: +45 4525 5170 Fax: +45 4596 3171

Email: lene@cti.dtu.dk

Keywords: Methodology for Societal Problems, Strategic Planning

Methodological approaches for solving, supporting or structuring complex decision making are usually selected given the existing knowledge and characterizations of the problem associated. This means that there is a tendency to apply almost predetermined approaches for specific problems. Over the years, this has sometimes created unforeseen 'failures' and a debate about whether applying hard or soft OR as alternatives for obtaining successful decision making. This paper takes its basis in a case study and discusses the need for application of a system of approaches in support for complex decision making. The case study refers to strategic decision making on Iceland in terms of the fish meal and oil industries. The Icelandic economy is highly dependent on these industries and with significant stock declines and new increasing global competition, a strategy has to be formed to cope with the complex situation. The study illustrates the need for using a system of soft (and hard) approaches to support the whole decision making process and to address the different levels of complexity arising.

University Utrecht, Department of Economics

Faculty of Law, Kromme Nieuwgracht 22

3512 HH Utrecht, The Netherlands, Tel.:030-2537100

*Tel: 030-2537118, Email: Bahlmann@planet.nl

Keywords: Complex Societal Problems, Disabled People

Compared to its neighbouring countries the Netherlands has a relatively large number of people whom receive income support under the various Sicknesses and Disability Insurance Acts. Because of the enormous costs involved, a considerable number of measures have been taken with a view to reducing this large number of people who live on benefits instead of on wages.

Although these measures are fairly diverse, it is possible to make a broad distinction between two types of measures. Roughly speaking, there are measures, which primarily aim at producing a reduction the number of people who live from the benefits and in the expenditures on benefits, and there are measures, which are intended to bring about reintegration of beneficiaries in the workforce.

Statistics show nevertheless, that in spite of all this the rehabilitation or reintegration of partially disabled people in the workforce still proceeds very slowly. Over the last few years the problem of insufficient reintegration of partially disabled people has become an issue of growing political concern. The increasing awareness of the great importance of having as many as possible people participate in the labour process plays a significant role. Sprouting from this awareness, the promotion of reintegration of partially disabled people was given highest priority during the Nineties. In order to achieve this goal, this time emphasis was put on strengthening the sense of individual responsibility of employers and employees to co-operate in ensuring the financial feasibility of the social security system. Active reintegration policy at company level seems an effective way to meet this responsibility.

But what factors contribute to successful reintegration of partially disabled into the labour process at company level including the effects of regulations in as far as this concerns the situation within the company? This seems to be a complicated process, in which several factors on macro, meso and micro level have their influences. The organisation forms part of the environment. The government, local authorities and social partners all influence the employer's behaviour concerning reintegration of partially disabled employees. One mustn't look at the employer as a person of a group, since several key figures inside the organisation like management staff, the personnel functionary, direct chief and the company doctor all influence the reintegration process.

From this the question arises: What are these critical success factors exactly for key figures leading to a successful reintegration of partially disabled people?

One way to find out is doing several best practise case studies at six Dutch companies and map out this complex process.

Weick's theory is useful to analyse this process, since the theory doesn't take the view of ones individual behaviour, but emphasises the interaction patron, which restricts to the behaviour of the key figures' relative position. These interaction processes together form a process, at which there is a distinction between negative and positive feedback mechanisms. By doing case studies and applying Weicks' theory, one tries to contribute to the insight regarding the complexes which play a part by the decision making process. This process precedes to the recruiting and keeping of partially disabled employees.

In our paper we shall elaborate on the outcome of the six case studies in The Netherlands. We will also discuss the policy in Germany and Sweden.

Department of Management Information Systems,

University College Dublin, Belfield, Dublin 4, Ireland.

Email: Cathal.brugha@ucd.ie

*Department of Mathematics, Royal Holloway, University of London,

Egham Hill, Egham, Surry TW20 0EX, England.

Keywords: Complex Issues

In the course of meeting at several conferences the authors discovered parallels between Brugha's Eight Adjustment Types in Nomology and Bowen's Eight Faces of Research. This paper explores these commonalties to determine if there is sufficient evidence to suggest that they reflect an inherent common structure.

University of Asia and the Pacific

Institute for Environmental Studies, College of Arts and Sciences

Pearl Drive, Ortigas Centre, Pasig city 1605, MetroManila, Philippenes

Tel: +632-634-2829, Fax: +632-634-2827, Email: theta@cas.uap.edu.ph

Keywords: Operational Research, Complex Societal Problems

This article considers the nature, historical background, basic elements and stages of operations research. Several examples on the application of operations research to

specific scientific, environmental and business problems are considered to illustrate the methodology.

University of Ljubljana, Faculty of Electrical Engineering, Ljubljana, Slovenia

Email: Marjan.Vezjak@fe.uni-lj.si

* Agencija za Radioaktivne Odpadke, Ljubljana, Slovenia

** Ministry of Defence, Ljubljana, Slovenia

Keywords: Multi Criteria Decision Making, Nuclear Waste

Decision-making is permanent issue in all spheres of activities. As a part of daily routine ease, intuition or plain reason can perform it. The complexity of the given problem rises as more alternative and parameters arise as a consequence of the analysis of the complex interdisciplinary societal problem. The aim of location selection for low and medium radioactive nuclear waste disposal (which represent in Slovenia a complex societal problem) was to show the applicability of PROMETHEE DSS method. The results have proven the method appropriate as a decision making tool for this purpose.

Session 8:5: chair Prof. Dr. Heiner Müller-Merbach

Technical University of München, Connollystr. 32, D-80809 München, Germany

Tel: 089 289-24521 Fax: 089 289-24535, Email: k.h.leist@sppae.wiso.tu-muenchen.de

Keywords: Personal Interests, Self-Organization

Abstract:

The growth of research on personal interests in the last decade indicates the continuing importance of this pedagogical and psychological construct motivating individual involvement with long-term goals. The presentation will begin by briefly reviewing the different perspectives of the contemporary theory and of empirical studies, which are relevant in the field of value-oriented behavior. Finally we will discuss the question how to bridge the gap between genuine interests, especially personal enjoyment and self-direction, and the moral issue in collective behavior and action. Challenging approaches to study this problem may be drawn form the theories of self-organization and synergetics because they deal which systems composed of many parts which, by their corporation, can produce patterns of coherent action on macroscopic scales.

Department of Methodology and Statistics, Faculty of Social Science, Utrecht University

Heidelberglaan 2, 3584 CS Utrecht, The Netherlands

Tel. work: +31 30 2534911, Fax: +31 30 2535797, Tel. home: +31 20 6622002

Email: c.vandijkum@fss.uu.nl, http://www.fsw.ruu.nl/ms/cvd/cvd.htm

Keywords: Knowledge-Based Systems, Simulation, Managing Uncertainty

In the next millennium society will confront us with problems that are very difficult to handle. In the natural sciences as well in the social sciences there are some pioneers who try to handle these complex societal problems in an interdisciplinary way. In operational research it is for example a good tradition to use the best of available techniques not only to solve technical problems, but also to try to solve societal problems. In the social sciences on the other hand some researchers try to combine the best of techniques of the social sciences with methods from the field of operational research to gain progress in the handling of these problems.

However concerning the question which approaches are adequate to handle complex societal problems and are leading to valid knowledge many answers are not yet given.

In this paper it is tried to give some answers to this challenging question.

The first question answered in the paper is: what are complex societal problems? With the pioneering work of the members of the IFORS group Methodology for Complex Societal Problems (for example: DeTombe 1994, DeTombe & Dijkum 1996) as a starting point the concept of complex societal problem is defined. Thereby it is argued that the mathematical theory of complexity is needed to define the concept of complex (societal) problems in an adequate way (Dijkum 2000, forthcoming).

In this interdisciplinary setting the paper is focused on the second question: how can the mathematical theory of complexity help to model complex (societal) problems in such a way that it leads to valid knowledge? Especially the method of computer simulation is introduced to demonstrate what complexity means, also in the social domain, and how non-linear models of phenomena can help to get insight in complex (societal) problems. The demonstration is illustrated by a number of research examples. The question how in such a way valid (scientific) knowledge is developed is answered from these examples in a global way (Dijkum & DeTombe 1998). An example of the practice of social research from the author and colleagues (Kuijk 1996) is taken to answer this question in a more systematic and specific way.

Dr. Dorien J. DeTombe, Chair Euro Group Complex Societal Problems, Delft University of Technology

School of Technoly, Policy and Management

Chair Operational Research Euro Working Group Complex Societal Problems

P.O. Box. 3286, 1001 AB Amsterdam, The Netherlands, Europe

Tel: +31 20 6927526

E-Mail: DeTombe@lri.jur.uva.nl

http://www.geocities.com/doriendetombe

Keywords: Method, Complex Societal Problems, Evaluation

Testing methods for complex real life problems is a complex task in itself. Usual methods are tested in a controlled environment in which participants and variables are more or less controlled. Using the method meant for real life problem handling, in which real experts and actors play a role, in an artificial environment, does not give the right view on the acting of the method. Testing a method in a real life situation makes controlling and a comparison of the participants, the problem, the variables and sometimes even of the selected hypotheses almost impossible. However to prevent this dilemma, there is a possibility to make a logical test: a development test of the method, by using the method in a real life situation and see whether the tasks, the description and the goals of the method are logical, clear and complete. Based on this test the method can be tested in real life. The real life situation should be selected in view of the goals of the method. Because it is not possible to compare two the same cases which each other the methods should be validated which its own goals land hypotheses. An example is given of the Compram method (DeTombe 1994).

Delft University of Technology

School of Systems Engineering, Policy Analysis and Management

P.O.Box 5015, 2600 GA Delft, The Netherlands

Tel.: +31 15 278 8071, Fax: +31 15 278 3422, Email: berte@sepa.tudelft.nl

http://www.sepa.tudelft.nl/webstaf/enserink/index.htm

Keywords: Forecasting, Managing Uncertainty

Problem formulation is one of the hardest and most neglected phases of problem solving. A problem can be defined as a gap between a normative standard and the existing or desired future situation. The desired situation may be known, but the future is inevitably shrouded in uncertainty. Despite the variety of 'clever' techniques at our disposal looking into the future is difficult.

The subject of this paper is coping with this kind of uncertainty in problem formulation. As pro-active policy making and anticipating the unexpected is required, scenario analysis might serve as a sensitivity analysis of the problem formulation. The paper describes the application of a method for designing strategic scenarios for problem formulation. The method is developed and tested in a policy analysis teaching environment.

University of Natal, Pietermaritzburg, Private bag X1, School of Maths,Stats & IT,

3201, Scottsville, South Africa

Tel. 2733-2605645, Fax 2733-2605648, Email: opetkova@it.unp.ac.za

Keywords: Societal Problems/Systems, Multi-criteria Decision Analysis, Decision Technology

A framework for solving complex problems based partly on Multimethodology and other recent advances in Systems Thinking is discussed. It includes techniques from MCDA, Soft Systems Thinking and Critical Systems Heuristics. It was applied by the author to problems of organizational transformation, evaluation of factors affecting software development productivity in an IT organization and elsewhere. The possible mechanism of interaction between the techniques as well as the issue of paradigm incommensurability links ideas by G. Midgley, M. Jackson and J. Hassard. The resulting framework is combining the strong aspects of the techniques involved. It is applicable to a wide range of complex societal problems requiring both soft and harder approaches.

Shahriari, Mohammad, Mattos, Luis de^* & Örtengren, Roland^**

Department of Ergonomics, Chalmers University of Technology

SE-412 96 Gothenburg, Sweden, Phone: +46 31 772 3639, Fax: +46 31 772 3660,

Email: msh@mvd.chalmers.se, ^*Email: mattos@mvd.chalmers.se

8:7:1 OR/MS Process: Parallel Components versus Consecutive Phases

Müller-Merbach, Heiner
Universitat Kaiserslautern, P.O.Box 3049  D-67653 Kaiserslautern, Germany
Tel: +49-631-205-2982, Fax:+49-631-205-338, Email:hmm@sozwi.uni-kl.de

Abstract: