TeCS

TeCS数据集由上海交通大学创建，专注于机器翻译中的时态一致性问题。该数据集包含552对法语-英语平行语句，每条语句都标注了时态信息，用于评估和改进机器翻译系统在处理时态预测方面的性能。数据集的创建过程涉及从广泛使用的Europarl数据集中提取并筛选具有复杂语法结构的语句，确保数据集的代表性和难度。TeCS数据集的应用领域主要集中在提高机器翻译系统的时态预测准确性，解决跨语言交流中的时态误解问题。

The TeCS dataset was developed by Shanghai Jiao Tong University, focusing on the issue of tense consistency in machine translation. It contains 552 French-English parallel sentence pairs, each annotated with tense information, which is used to evaluate and improve the performance of machine translation systems in handling tense prediction. The construction process of the dataset involves extracting and filtering sentences with complex syntactic structures from the widely used Europarl dataset, to ensure the representativeness and difficulty of the dataset. The main application scenarios of the TeCS dataset focus on enhancing the accuracy of tense prediction for machine translation systems and solving the problem of tense misunderstanding in cross-linguistic communication.