Representing Community Concerns in Agent-Based Models: a Web 2.0 Approach

Decision-makers in an online deliberative forum require access to tools to help them understand the potential impact of proposed policies. Agent-based models can help to predict such impacts. However, the questions being asked by many existing models are not always relevant to the concerns of communities affected by proposed policies. In this exploratory paper, we outline a methodology for exploiting Web 2.0 to capture the concerns of diverse communities and translating these concerns into formalisms that can generate agent-based simulations. We argue that this has the potential to support collaboration and consensus formation in addition to improving the relevance of the models. However, it is important that such a participatory approach can be reconciled with scientific development of models, based on empirical evidence.