Local governments face increasingly complex decisions and must inevitably rely on professional staff with specialised knowledge. However, ordinary citizens and stakeholders are demanding the right to directly participate in governmental decisions. What is the appropriate division of labour? The article proposes a practical approach to participatory decision making that tries to combine administrative efficiency and democratic legitimacy. The approach decomposes the decision problem into a number of discrete stages. An action research methodology is used to illustrate the application of the method. Specifically, we decompose a roaming horse' problem in the interior of British Columbia to identify the information requirements for each stage of the model. We use a series of web-based Delphi surveys to elicit specific information from citizens regarding objectives and potential alternatives. The survey results suggest that a relatively simple and cost-effective deliberative tool like Delphi can facilitate an effective division of labour between citizens and government experts.

This chapter develops a model of open source disruption in enterprise software markets. It addresses the question: Is free and open source software (FOSS) likely to disrupt markets for enterprise business applications? The conventional wisdom is that open source provision works best for low-level systemoriented technologies while large, complex enterprise business applications are best provided by commercial software vendors. The authors challenge the conventional wisdom by developing a two-stage model of open source disruption in business application markets that emphasizes a virtuous cycle of adoption and lead-user improvement of the software. The two stages are an initial incubation stage (the I-Stage) and a subsequent snowball stage (the S-Stage). Case studies of several FOSS projects demonstrate the model’s ex post predictive value. The authors then apply the model to SugarCRM, an emerging open source CRM application, to make ex ante predictions regarding its potential to disrupt commercial CRM incumbents.

The objective of this paper is to use a challenging real-world problem to illustrate how a probabilistic predictive model can provide the foundation for decision-analytic feedforward control. Commercial data mining software and sales data from a market research firm are used to create a predictive model of market success in the video game industry. A procedure is then described for transforming the classification trees into a decision-analytic model that can be solved to produce a value-maximizing game development policy. The video game example shows how the compact predictive models created by data mining algorithms can help to make decision-analytic feedforward control feasible, even for large, complex problems. However, the example also highlights the bounds placed on the practicality of the approach due to combinatorial explosions in the number of contingencies that have to be modeled. We show, for example, how the "option value" of sequels creates complexity that is effectively impossible to address using conventional decision analysis tools.

This article develops a model of open source disruption in enterprise software markets. It addresses the question: Is free and open source software (FOSS) likely to disrupt markets for commercial enterprise software? The conventional wisdom is that open source provision works best for low-level system-oriented technologies, while large, complex enterprise business applications are best served by commercial software vendors. The authors challenge the conventional wisdom by developing a two-stage model of open source disruption in enterprise software markets that emphasizes a virtuous cycle of adoption and leaduser improvement of the software. The two stages are an initial incubation stage (the I-Stage) and a subsequent snowball stage (the S-Stage). Case studies of several FOSS projects demonstrate the model’s ex post predictive value. The authors then apply the model to SugarCRM, an emerging open source CRM application, to make ex ante predictions regarding its potential to disrupt commercial CRM incumbents.

Data-mining technologies are within the grasp of many organizations. Commercially available data-mining packages make it relatively easy for firms to transform their data resources into predictive models. Yet, despite technological advances, the precise manner in which data-mining output should be incorporated into an organization's decision-making processes remains unclear. This paper attempts to clarify the role of data mining by situating it within the context of Simon's model of decision making. We use a complex decision problem from the video game development industry to illustrate several practical challenges managers face when using data-mining output as a decision making input. We then show how some of these challenges can be overcome by incorporating data-mined predictive models into a conventional decision-analytic formulation of the problem.

In writing this paper, our objective was to use the concept of internal market failure to explain why many knowledge management initiatives fall short of expectations. We re-examined the convention...

Is open source software likely to disrupt the commercial customer relationship management (CRM) market? To address this question, we develop a dynamic model of disruption based on two complementary perspectives from the adoption literature: Diffusion of Innovations and Economics of Technology Standards. We use case studies of several well-known FOSS projects to illustrate the model and demonstrate the model's ex post predictive value. We then apply the model to SugarCRM, an emerging open source customer relationship management (CRM) application, to make ex ante predictions regarding its potential to disrupt commercial CRM incumbents.

The purpose of this research is to examine whether decision-theoretic planning techniques can be used to help managers evaluate strategic options in complex and uncertain environments. Firms faced with choices such as whether to acquire a start-up, develop a new product, or invest in updated production technology continue to make decisions based on unreliable heuristics, “gut feel” or misleading financial measures such as net present value (NPV). In this paper we show that decision-theoretic planning techniques originally developed for robot planning permit us to gain the insights provided by real options analysis without working within the restrictions of models designed to price financial options or incurring the overhead of constructing huge decision trees. A biotechnology licensing problem similar to those addressed elsewhere in the real options literature is used to illustrate the methodology and demonstrate its feasibility.

The primary advantage of using simulated internal markets to solve complex resource allocation problems is that markets permit much of the computation of a solution to be distributed over a large number independent agents running on separate processors. The difficulty that arises in the context of NP-hard resource allocation problems is that the market for resources inevitably takes the form of a combinatorial auction, which induces a different type of NP-hard problem. We examine an important class of stochastic, intrafirm resource allocation problems and ask whether economic constructs, such as agents, markets, and prices, provide a useful foundation for structuring decentralized heuristic solution techniques. We show how complex exchange protocols can help market-based search techniques avoid the local maxima problems associated with other greedy search heuristics and converge on good equilibrium solutions.

In writing this paper, our objective was to use the concept of internal market failure to explain why many knowledge management initiatives fall short of expectations. We re-examined the conventional view of knowledge as a pure public good and developed a typology of knowledge as a heterogeneous public good. This permitted us to identify the different sources of internal market failure that impeded knowledge creation and sharing within firms. We then analyzed generic managerial responses to internal market failure and showed how the effectiveness of each response was limited by the nature of knowledge as a tradable commodity. We concluded by presenting a preliminary framework for knowledge management based on the enforcement of dynamic internal property rights. The objective of a dynamic response to internal knowledge market failure was seen as an attempt to balance individual incentives with the need to create and share knowledge throughout the organizational.

This paper examines a number of theoretical and practical issues concerning the use of decision-theoretic planning to implement agents in market-based systems. The markets considered here result from the decomposition of complex, intra-organizationai resource allocation problems such as manufacturing scheduling. Although these problems can be formulated as monolithic optimization problems, they tend to be much too large to solve in practice. Markets provide a means of decomposing large resource allocation problems and distributing the computation of a solution over many processors. An important precondition of efficient markets is agent-level rationality. Decision theoretic planning can be used to implement economic rationality and thus decision theoretic planning agents fit well into market-based approaches. The primary challenge in building decision theoretic planning agents is the size of the agent’s state space. Although marketbased decomposition results in agent-level problems that are much smaller than the original resource allocation problem, the price mechanism used to achieve independence of the agent-level problems requires that the agents plan over a large number of different resource contingencies. This requirement exacerbates the state space explosion that characterizes decision-theoretic planning. The application of state space reduction techniques such as structured dynamic programming and reachability analysis are shown to yield significant reductions in the effective size of the agent-level problems and thereby increase the applicability of decision theoretic planning techniques in market-based systems.

In this thesis, a framework for market-based resource allocation in manufacturing is developed and described. The most salient feature of the proposed framework is that it builds on a foundation of well-established economic theory and uses the theory to guide both the agent and market design. There are two motivations for introducing the added complexity of the market metaphor into a decision-making environment that is traditionally addressed using monolithic, centralized techniques. First, markets are composed of autonomous, self-interested agents with well defined boundaries, capabilities, and knowledge. By decomposing a large, complex decision problem along these lines, the task of formulating the problem and identifying its many conflicting objectives is simplified. Second, markets provide a means of encapsulating the many interdependencies between agents into a single mechanism—price. By ignoring the desires and objectives of all other agents and selfishly maximizing their own expected utility over a set of prices, the agents achieve a high degree of independence from one another. Thus, the market provides a means of achieving distributed computation. To test the basic feasibility of the market-based approach, a prototype system is used to generate solutions to small instances of a very general class of manufacturing scheduling problems. The agents in the system bid in competition with other agents to secure contracts for scarce production resources. In order to accurately model the complexity and uncertainty of the manufacturing environment, agents are implemented as decision-theoretic planners. By using dynamic programming, the agents can determine their optimal course of action given their resource requirements. Although each agent-level planning problem (like the global level planning problem) induces an unsolvably large Markov Decision Problem, the structured dynamic programming algorithm exploits sources of independence within the problem and is shown to greatly increase the size of problems that can be solved in practice. In the final stage of the framework, an auction is used to determine the ultimate allocation of resource bundles to parts. Although the resulting combinational auctions are generally intractable, highly optimized algorithms do exist for finding efficient equilibria. In this thesis, a heuristic auction protocol is introduced and is shown to be capable of eliminating common modes of market failure in combinational auctions.

The overall objective of this program of research is to develop a model of Internet-induced channel competition. In this paper, we focus on the ways in which retail channel technology—specifically, the online vs. bricks and mortar stores—affects the feasible trade-offs that firms can make between price and desirable attributes of their product/service bundles. This paper treats products as a bundle of the physical good and the fulfillment or transaction technology, and proposes a model of competition in the price-attribute space to illustrate the tradeoffs for consumers and producers. This model is grounded in demand, production, and hedonic theory, and relates the attributes (or “quality”) of products to their observed prices. Our objectives in future research are to refine the analytical model and to find evidence that (1) the functional forms assumed in the model are consistent with the price/attribute trade-offs observed in practice and (2) the observed competitive responses of firms dominated by online competitors are consistent with those prescribed by our model.

The initial goal of this project was to provide Alcan Rolled Products Company of Kingston, Ontario, with control charts to monitor its new Paint Line Three (PL3) aluminium coil coating facility. As the project progressed, however, it became apparent that the following factors made PL3 ill-suited to traditional statistical quality control methods: the inexact nature of the coil coating science, the level of process interdependence and complexity of the operation, and the economic and technical infeasibility of installing sophisticated measuring devices at all stages of the coating line.Consequently, an effort was made to model the entire system on a macroscopic, rather than a microscopic, level and to use the results of qualitative lab tests to make inferences about the paint line's process capability. To accomplish this, the length of scrap produced by every lot processed in a 1-year period was used to create a statistical model of the of the frequency and severity of the two most costly quality shortcomings: surface defects and coating problems.This article outlines the methodology used and demonstrates how a complex phenomenon such as scrap creation can be modeled by using a combination of well-known statistical distributions. The statistical model can be used to transform raw data into meaningful graphical representations of large-scale system performance.