Geocomputing Laboratory, Department of Geomatics Engineering, University of Calgary, 2500 University Drive NW, Calgary, AB, Canada T2N 1N4.
J Environ Manage. 2013 Nov 15;129:309-23. doi: 10.1016/j.jenvman.2013.07.028. Epub 2013 Aug 23.
Decision making in land management can be greatly enhanced if the perspectives of concerned stakeholders are taken into consideration. This often implies negotiation in order to reach an agreement based on the examination of multiple alternatives. This paper describes a spatial web/agent-based modeling system that was developed to support the negotiation process of stakeholders regarding land development in southern Alberta, Canada. This system integrates a fuzzy analytic hierarchy procedure within an agent-based model in an interactive visualization environment provided through a web interface to facilitate the learning and negotiation of the stakeholders. In the pre-negotiation phase, the stakeholders compare their evaluation criteria using linguistic expressions. Due to the uncertainty and fuzzy nature of such comparisons, a fuzzy Analytic Hierarchy Process is then used to prioritize the criteria. The negotiation starts by a development plan being submitted by a user (stakeholder) through the web interface. An agent called the proposer, which represents the proposer of the plan, receives this plan and starts negotiating with all other agents. The negotiation is conducted in a step-wise manner where the agents change their attitudes by assigning a new set of weights to their criteria. If an agreement is not achieved, a new location for development is proposed by the proposer agent. This process is repeated until a location is found that satisfies all agents to a certain predefined degree. To evaluate the performance of the model, the negotiation was simulated with four agents, one of which being the proposer agent, using two hypothetical development plans. The first plan was selected randomly; the other one was chosen in an area that is of high importance to one of the agents. While the agents managed to achieve an agreement about the location of the land development after three rounds of negotiation in the first scenario, seven rounds were required in the second scenario. The proposed web/agent-based model facilitates the interaction and learning among stakeholders when facing multiple alternatives.
如果考虑到相关利益相关者的观点，土地管理中的决策可以得到极大的改善。这通常意味着需要进行协商，以便根据对多个备选方案的审查达成协议。本文描述了一个空间网络/基于代理的建模系统，该系统旨在支持加拿大艾伯塔省南部土地开发的利益相关者的协商过程。该系统在基于代理的模型中集成了模糊层次分析程序，并在通过 Web 界面提供的交互式可视化环境中，以促进利益相关者的学习和协商。在预协商阶段，利益相关者使用语言表达来比较他们的评估标准。由于这种比较的不确定性和模糊性，因此随后使用模糊层次分析法来对标准进行优先级排序。协商始于用户（利益相关者）通过 Web 界面提交的开发计划。一个名为提议者的代理代表计划的提议者接收此计划，并开始与所有其他代理进行协商。协商是逐步进行的，代理通过为其标准分配一组新的权重来改变其态度。如果未达成协议，则提议者代理会提出新的开发地点。这个过程会重复，直到找到一个满足所有代理的某个预定义程度的地点为止。为了评估模型的性能，使用四个代理（其中一个是提议者代理）模拟了协商过程，使用了两个假设的开发计划。第一个计划是随机选择的；另一个计划是在一个对一个代理非常重要的区域选择的。虽然在第一种情况下，代理在三轮协商后就成功就土地开发的地点达成了协议，但在第二种情况下则需要七轮协商。提出的基于网络/代理的模型有助于在面对多个备选方案时，利益相关者之间的互动和学习。
