SBIR-STTR Award

A project to explore extending electronic prediction markets for use as a decision aid tool. Investigation focuses on issues of a military interest but has broader implications. Issues explored include methods to identify suitable issues, relevant events, screen events for suitability for electronic markets, identification of market participants, and identification of appropriate incentives. Related issues of legality, acceptance, regulatory restrictions, and adjunct decision making benefits are also explored. Test markets are designed and demonstrations planned. Improved accuracy of predictions resulting from aggregation of individual knowledgeable inputs, even when knowledge is of limited scope. Responsive nature of predictions offered to users with minimum intrusion on responding individuals. Anonymity and potential for personal reward provides envrionment for honest response. Methodology offers improvements over surveys.

A project to explore extending electronic prediction markets for use as a decision aid tool. Investigation focuses on issues of a military interest but has broader implications. Issues explored include methods to identify suitable issues, relevant events, screen events for suitability for electronic markets, identification of market participants, and identification of appropriate incentives. Test markets are designed and demonstrations planned. Demonstration applications are planned in anticipation of using results in quasi operationa decision making settings. At least two such demonstrations are planned for full implementation through analsysis of effectivity. Additional demonstrations will be explored and begun. Supporting laboratory experiments and field demonstrations will be employed to further the theory of application to different decision making conditions. Tools to enhance the operation and ease the application to different settings will be developed.

Keywords:
Market, Prognostics, Decision Support, Risk Assessment, Bayesian Network Models, Electronic Markets, Forecasting, Consensus Aiding Groupware