wangzailiunai/XQ-MEval
收藏资源简介:
XQ-MEval是一个用于评估自动评估指标在跨语言评分偏差方面的基准数据集,基于[CC BY-S 4.0](https://creativecommons.org/licenses/by-sa/4.0/)许可发布。该数据集通过在高质量翻译中注入多维质量度量(MQM)定义的不同数量的错误,实现了跨语言的可控和可比较的翻译质量。数据集包含多种语言对(如英语-中文、英语-老挝等)和错误类型(如添加、遗漏、误译等),并提供了详细的错误分布统计。数据集的构建基于Flores+数据集,并提供了灵活的构建流程,适用于不同语言和错误类型的扩展。
XQ-MEval is a benchmark released under [CC BY-S 4.0](https://creativecommons.org/licenses/by-sa/4.0/) for evaluating automatic metrics with respect to cross-lingual scoring bias. This dataset is constructed by injecting varying numbers of Multidimensional Quality Metric (MQM)-defined errors into high-quality translations, enabling controlled and comparable translation quality across languages. It includes multiple language pairs (e.g., English-Chinese, English-Lao) and error types (e.g., Addition, Omission, Mistranslation), with detailed error distribution statistics. The dataset is based on the Flores+ dataset and features a flexible construction pipeline adaptable to different languages and error types.





