An Optimal Algorithm for Raw Idea Selection under Uncertainty

At the first gate of an innovation process, a large number of raw ideas must be evaluated and those good enough to continue to the next phase be selected. No information about these ideas is available, so they have a high level of uncertainty. We present an algorithm that selects and ranks a set of alternatives in optimal time. The algorithm addresses uncertainty by allowing decision-makers to specify missing information that affect the outcome of their judgments. It generates multiple partial rankings efficiently according to the various possible combinations of missing items of information and identifies the set of items that are needed to obtain a unique result. In this manner, we can reduce the uncertainty in the selection procedure and make explicit expert knowledge that is relevant to the evaluation process. The algorithm is intended for use in a collaborative tool for corporations who utilize a structured innovation process.