SEAGraph: Unveiling the Whole Story of Paper Review Comments

TL;DR

SEAGraph tackles the challenge of understanding reviewer feedback by introducing a dual-graph framework that captures both the paper's internal semantics and its broader domain context. It builds a semantic mind graph $\,\mathcal{G}_S\$ and a hierarchical background graph $\,\mathcal{G}_H\$, then performs retrieval over these structures to produce coherent explanations for each review comment, which are fed to LLMs for final reasoning. The method combines chunk-level semantic modeling with a three-layer background knowledge graph and demonstrates strong performance in both automated and human evaluations, outperforming DirectInfer and naive RAG baselines across relevance, clarity, novelty, and persuasiveness. This approach has practical implications for making peer review more transparent and efficient by aligning authors’ revisions with reviewer intent through structured, evidence-backed explanations.

Abstract

Peer review, as a cornerstone of scientific research, ensures the integrity and quality of scholarly work by providing authors with objective feedback for refinement. However, in the traditional peer review process, authors often receive vague or insufficiently detailed feedback, which provides limited assistance and leads to a more time-consuming review cycle. If authors can identify some specific weaknesses in their paper, they can not only address the reviewer's concerns but also improve their work. This raises the critical question of how to enhance authors' comprehension of review comments. In this paper, we present SEAGraph, a novel framework developed to clarify review comments by uncovering the underlying intentions behind them. We construct two types of graphs for each paper: the semantic mind graph, which captures the authors' thought process, and the hierarchical background graph, which delineates the research domains related to the paper. A retrieval method is then designed to extract relevant content from both graphs, facilitating coherent explanations for the review comments. Extensive experiments show that SEAGraph excels in review comment understanding tasks, offering significant benefits to authors. By bridging the gap between reviewers' critiques and authors' comprehension, SEAGraph contributes to a more efficient, transparent and collaborative scientific publishing ecosystem.

Figures (10)

• Figure 1: SEAGraph can help authors better understand reviewers’ comments by providing detailed insights and evidence.
• Figure 2: The overall framework of SEAGraph consists of the construction of the semantic mind graph and the hierarchical background graph, along with the corresponding retrieval module. The final retrieved content is fed into LLMs for review comment understanding.
• Figure 3: Chunks from different sections can form a coherent logic chain.
• Figure 4: Research paper topic distribution across six key areas. NLP: Natural Language Processing; MM: Multimodal Learning; CV: Computer Vision; RLearn: Representation Learning; Theory&Opt.: Theory and Optimization; Robust.: Robustness.
• Figure 5: Distribution of Review Lengths.