Bayesian-based systems have been criticised in the past, because users perceived them as opaque, because they did not find the reasoning intuitive and because unguided cognition and behaviour in circumstances of uncertainty seldom conforms to the Bayesian model. However, advances in artificial intelligence based explanation systems together with the pervasive access to the Internet now offer ways in which these techniques can be used to inform the decision making process in many more ways, encouraging constructive discussion and debate between the many parties of the process. Complex decisions typically involve a range of different stakeholder groups, including the general public, each holding different beliefs and values. A failure to consult with these groups and take account of differences in beliefs and values can severely limit the effectiveness of a decision both in terms of what is chosen and how acceptable it will be to key stakeholders. Thus, such methodologies provide a route for delivering the benefits promised to the Information Society and e-government, and a step towards true e-democracy.
The programme will address the many challenging research problems in implementing and evaluating the benefits of such a development, by introducing methodology in the five stages of decision support (problem structuring, belief modelling, preference modelling, optimisation, sensitivity analysis) and promoting a decision support and communication infrastructure (DS/CI), which entails risk communication issues. Our development will be fundamentally multi-disciplinary, drawing on skills from world class groups across Europe with expertise in, inter alia, artificial intelligence, behavioural science, computer science, cognitive science, decision analysis, mathematics, operational research and statistics. The methodology and the DS/CI will promote interchange of information and value judgements between all parties to a major decision. Decision makers will be able to consult and communicate with stakeholders throughout the decision making process not just by reports, written submissions, and opinion polling, but interactively via web-based technologies which explore their various perspectives within a common and inclusive decision model. The intention is to focus on the key issues and not be diverted by entrenched positions or worldviews that are less significant in the context of the decision.