dialect_eval
收藏资源简介:
dialect_eval数据集由苏黎世大学创建,包含1997个句子,专门设计用于评估机器翻译指标在处理非标准化方言时的性能,特别是英语到两种瑞士德语方言的翻译。该数据集包括人工翻译和人工评价,用于自动机器翻译输出。此外,还创建了一个挑战集,用于测试方言变异下的指标性能。数据集的应用领域主要集中在提高机器翻译系统在处理方言和语言变体时的准确性和鲁棒性。
The dialect_eval dataset was created by the University of Zurich. It comprises 1997 sentences and is specifically designed to evaluate the performance of machine translation metrics when handling non-standardized dialects, particularly for translation tasks from English to two Swiss German dialects. This dataset includes human translations and human evaluations for automatic machine translation outputs. Additionally, a challenge set has been developed to test the performance of translation metrics under dialectal variation. The main application of this dataset focuses on enhancing the accuracy and robustness of machine translation systems when processing dialects and language variants.




