TheMrguiller/semantic-consistency-annotations
收藏资源简介:
该数据集专注于句子转换任务,包含原始句子及其标签、转换后的句子、转换类型定义,以及多个评估指标(如转换正确性、意义保留、标签保留、可转换性等),这些指标可能来自不同评估者或系统(通过UUID标识)。数据集用于训练或评估自然语言处理模型在句子转换中的性能,例如语义保持或标签一致性分析。
This dataset focuses on the sentence transformation task. It includes original sentences with their corresponding labels, transformed sentences, definitions of transformation types, and multiple evaluation metrics, such as transformation correctness, meaning preservation, label preservation, convertibility, and so on. These metrics may originate from different evaluators or systems, which are identified by UUIDs. This dataset is intended for training or evaluating the performance of natural language processing (NLP) models on sentence transformation tasks, for example, semantic preservation or label consistency analysis.




