inno-idea-eval

Multi-persona research idea evaluation assigns three independent reviewer personas to assess a proposal across five dimensions. A final Area Chair meta-review aggregates the scores, resolves disagreements, and outputs a final decision based on configured thresholds.

You can evaluate research idea novelty and feasibility by running the idea through three independent reviewer personas. Each persona scores Clarity, Novelty, Validity, Feasibility, and Significance, providing structured scores and narrative rationale for each dimension.

A meta-review process consolidates multiple independent persona evaluations into a single final decision. An Area Chair aggregates individual scores across five dimensions, resolves disagreements between reviewers, and applies configured thresholds to determine the final outcome.

Yes, you can use structured prompts and references to ground research idea scores in evidence. The evaluation workflow requires prompts and references to drive evidence-based scoring, ensuring each persona's narrative rationale and structured scores are supported by provided materials.

A limitation of using independent persona reviews is that disagreements between the three reviewers must be resolved manually by an Area Chair meta-review. You must start three separate evaluator conversations, and the final decision reliability depends heavily on the quality of the provided references and prompts.