This repository contains a model designed to enhance the quality and fairness of the academic peer review process by identifying and resolving conflicts in reviewer feedback. The peer review process, critical to academic publishing, often encounters conflicting comments from different reviewers. This model aims to automatically detect and classify these conflicts, improving clarity, consistency, and efficiency in peer review analysis.
- Disagreement with Contradiction (Dc): Reviewers explicitly contradict each other’s conclusions or arguments.
- Disagreement without Contradiction (Dn): Reviewers diverge in opinion without direct contradictions.
- Agreement with Contradiction (Ac): Reviewers generally agree on the assessment but show contradictions on specific points.
The model analyzes reviewer feedback to detect these three primary types of conflicts, aiding editors and authors in efficiently resolving disagreements. This tool is intended to improve the transparency and fairness of peer review by promoting a more structured conflict-resolution process.
For a detailed overview, please refer to our ongoing paper edit link.