LawReasoningBenchmark
收藏资源简介:
本研究构建了一个高质量的法律案例描述数据集,其中包含多级别的标注事实、证据以及相关事实之间的链接和涉及的经验,遵循研究设计的模式,并展示了法律推理过程的明确路径。数据集通过众包方式构建,利用公开可用的数据,并实施严格的标注过程控制,确保了数据标注的可靠性和质量。该数据集旨在为法律推理模拟提供支持,解决法律领域中证据解释与司法决策之间的差距问题。
This study constructs a high-quality legal case description dataset that includes multi-level annotated facts, evidence, links between related facts, and relevant experience. Aligned with the pre-defined research design framework, the dataset showcases explicit pathways of the legal reasoning process. It is developed via crowdsourcing, leveraging publicly available data, and incorporates strict annotation process control to ensure the reliability and quality of the annotated data. This dataset aims to support legal reasoning simulations, bridging the gap between evidence interpretation and judicial decision-making in the legal domain.




