SLUE

SLUE数据集是由丰田技术学院芝加哥分校和ASAPP合作创建的，旨在为口语理解任务提供基准测试。该数据集包含5000个样本，主要用于评估自动语音识别（ASR）、命名实体识别（NER）和情感分析（SA）。数据集中的语音数据来自自然对话，确保了数据的真实性和多样性。创建过程中，研究团队对VoxCeleb和VoxPopuli数据集进行了转录和注释，以适应SLUE的需求。该数据集的应用领域广泛，包括但不限于语音识别技术的改进、情感分析算法的优化以及自然语言处理的研究。

The SLUE Dataset was co-created by Toyota Technological Institute at Chicago and ASAPP, with the aim of providing a benchmark for spoken language understanding tasks. It contains 5,000 samples, which are mainly used to evaluate automatic speech recognition (ASR), named entity recognition (NER) and sentiment analysis (SA). The speech data in the dataset is derived from natural dialogues, ensuring its authenticity and diversity. During the creation process, the research team transcribed and annotated the VoxCeleb and VoxPopuli datasets to meet the requirements of SLUE. This dataset has a wide range of application fields, including but not limited to the improvement of speech recognition technologies, the optimization of sentiment analysis algorithms and research in natural language processing.