CATSLU

该数据集名为CATSLU，包含了地图、音乐、视频和天气四个对话领域的内容。它被用于评估口语语言理解（SLU）系统，特别是在自动语音识别错误的环境下。该数据集的设计旨在为SLU系统在存在自动语音识别错误的情况下的性能提供一个基准。任务的焦点是口语语言理解。The dataset named CATSLU includes content across four dialogue domains: mapping, music, video, and weather. It is utilized to evaluate Spoken Language Understanding (SLU) systems, particularly in scenarios with automatic speech recognition (ASR) errors. This dataset is developed to establish a benchmark for SLU system performance under conditions involving automatic speech recognition errors. The task focuses on spoken language understanding.