KddRES

KddRES是由香港中文大学发布的第一个粤语知识驱动餐饮对话数据集，包含832个对话，源自10家风格各异的香港餐厅。该数据集设计了细粒度的槽和意图，以更好地捕捉语义信息，适用于构建定制化对话系统，特别适合中小企业如餐厅使用。数据集通过模拟真实场景中的餐厅信息查询和预订，详细记录了用户和系统之间的交互，旨在解决现有数据集在特定领域信息处理上的不足。

KddRES is the first Cantonese knowledge-driven catering dialogue dataset released by The Chinese University of Hong Kong. It contains 832 dialogues sourced from 10 Hong Kong restaurants with distinct styles. This dataset is equipped with fine-grained slots and intents to better capture semantic information, making it applicable for building customized dialogue systems, especially for small and medium-sized enterprises (SMEs) such as restaurants. By simulating real-world scenarios of restaurant information inquiry and reservation, it records detailed interactions between users and systems, aiming to address the shortcomings of existing datasets in specific domain information processing.