Civil Private Loan Disputes (PLD) Dataset

该数据集包含了超过30万份民间私人贷款纠纷的法院庭审记录，以及与之对应的判决结果。这些记录针对事实方面和实体一致性进行了注释，涉及12个事实方面和14种事实实体，且注释者之间具有高度的一致性（肯塔系数为0.9）。该数据集被划分为训练集、验证集和测试集三个子集。规模上，记录总数超过30万条，其中有45,531条案例进行了注释。这项任务的目的是进行对话摘要，同时关注事实不一致性的问题。

This dataset contains over 300,000 court trial transcripts of private civil loan disputes, paired with their corresponding judgment results. These records are annotated with respect to factual aspects and entity consistency, covering 12 factual aspects and 14 categories of factual entities, with high inter-annotator agreement (Cohen's Kappa score of 0.9). The dataset is split into three subsets: training set, validation set, and test set. In terms of scale, the total number of records exceeds 300,000, among which 45,531 cases have been annotated. The objective of this task is conversational summarization, with a focus on factual inconsistency.