COMPARE

COMPARE数据集由印度理工学院创建，专注于深度学习领域的研究论文同行评审中的比较讨论。该数据集包含117篇评审，涵盖约1,800个句子，旨在通过详细的分类体系分析比较讨论。数据集的创建涉及对大量评审句子的深入观察和分类，以及对每个句子的精细标注。COMPARE数据集的应用领域包括自动化识别和分析研究论文中的比较讨论，以辅助评审过程和提高论文评审的效率。

The COMPARE dataset was created by the Indian Institute of Technology, focusing on comparative discussions within peer reviews of research papers in the field of deep learning. Comprising 117 peer reviews totaling approximately 1,800 sentences, this dataset is designed to analyze comparative discussions via a detailed classification system. The construction of the COMPARE dataset involves in-depth observation and categorization of a large volume of review sentences, alongside fine-grained annotation for each individual sentence. Application areas of the COMPARE dataset include automated identification and analysis of comparative discussions in research papers, to support the peer review workflow and enhance the efficiency of paper reviewing.