VLSP 2025MLQA-TSR
收藏资源简介:
VLSP 2025MLQA-TSR 是一个关于越南交通标志法规的多模态法律问答数据集,旨在推动多模态法律文本处理的研究,并为构建和评估多模态法律领域的智能系统提供基准数据集。该数据集包含两个子任务:多模态法律检索和多模态法律问答。数据集由两个部分组成:法律检索和法律问答。法律检索子任务的目标是根据问题检索相关法律文档,而法律问答子任务的目标是根据问题和相关文档预测正确答案。该数据集由三个部分组成:训练集、公开测试集和私有测试集。此外,还提供了包含交通标志法规和道路交通秩序与安全法的法律数据库。
VLSP 2025 MLQA-TSR is a multimodal legal question answering dataset centered on Vietnamese traffic sign regulations. It aims to advance research in multimodal legal text processing and serve as a benchmark dataset for developing and evaluating intelligent systems in the multimodal legal field. This dataset includes two subtasks: multimodal legal retrieval and multimodal legal question answering, and consists of two core components: legal retrieval and legal question answering. The objective of the legal retrieval subtask is to retrieve relevant legal documents based on the input question, while the goal of the legal question answering subtask is to predict the correct answer using the given question and the retrieved relevant documents. The dataset is split into three subsets: training set, public test set, and private test set. Additionally, a legal database containing traffic sign regulations and the Law on Road Traffic Order and Safety is provided.
- 1通过越南信息科技大学, 胡志明市, 越南, 日本先进科学技术研究所, 石川县, 日本, 越南国立大学, 胡志明市, 越南 · 2025年



