DiP benchmark tests

DiP benchmark tests是由南洋理工大学和华沙理工大学联合创建的数据集，专注于评估机器翻译系统在处理语篇现象（如指代、词汇一致性、连贯性和可读性、语篇连接词翻译）方面的能力。该数据集通过自动数据提取方法构建，旨在为不同语言对的机器翻译模型提供挑战性的测试环境。数据集的应用领域主要集中在提高机器翻译系统的语篇处理能力，以期达到更接近人类翻译的质量。

The DiP benchmark test is a dataset jointly developed by Nanyang Technological University and Warsaw University of Technology, which focuses on evaluating the ability of machine translation systems to handle discourse phenomena, including reference resolution, lexical consistency, coherence, readability, and the translation of discourse connectives. Constructed through automatic data extraction methods, this dataset aims to provide a challenging test environment for machine translation models across various language pairs. The main application scenario of this dataset is to improve the discourse processing capabilities of machine translation systems, so as to achieve translation quality closer to that of human translators.