MAVEN

MAVEN是由清华大学开发的大规模通用领域事件检测数据集，包含4,480篇维基百科文档，118,732个事件提及实例，涵盖168种事件类型。该数据集旨在解决现有数据集数据稀缺和覆盖率低的问题，通过人工标注确保数据质量，适用于训练和评估现代神经网络方法。MAVEN的应用领域广泛，可用于提升事件检测模型的性能，尤其是在处理复杂和多样化的文本事件描述时。

MAVEN is a large-scale general-domain event detection dataset developed by Tsinghua University. It comprises 4,480 Wikipedia articles, 118,732 event mention instances, and covers 168 event types. This dataset aims to address the problems of data scarcity and low coverage in existing datasets, with data quality guaranteed through manual annotation, making it suitable for training and evaluating modern neural network methods. MAVEN has a wide range of application scenarios, and can be used to enhance the performance of event detection models, especially when processing complex and diverse textual event descriptions.