QSEM

QSEM数据集由京都大学和庆应大学合作创建，专注于评估疑问句的语义分析，特别是使用组合范畴语法（CCG）框架。该数据集包含138个样本，主要分为两类问题：复杂多样的形式语义问题和接近真实语言使用的问题。QSEM数据集的构建遵循了FraCaS的方法论，旨在为自然语言推理提供一个可靠的起点。此数据集不仅有助于理解范畴语法中的疑问句，还揭示了现有CCG解析器的挑战和局限性，为计算语言学及相关领域的未来研究和应用提供了宝贵的见解。

The QSEM dataset was collaboratively created by Kyoto University and Keio University, focusing on evaluating semantic analysis of interrogative sentences, particularly within the framework of Combinatory Categorial Grammar (CCG). It contains 138 samples, which are mainly divided into two categories of questions: complex and diverse formal semantic questions, and questions that approximate real-world language usage. The construction of the QSEM dataset follows the methodology of FraCaS, aiming to provide a reliable starting point for natural language inference. This dataset not only facilitates the understanding of interrogative sentences in categorial grammar, but also reveals the challenges and limitations of existing CCG parsers, offering valuable insights for future research and applications in computational linguistics and related fields.