PET
收藏资源简介:
PET数据集是由Bruno Kessler基金会和合作伙伴共同创建,专注于从自然语言文本中提取业务流程模型。该数据集包含45个文本描述,每个描述都经过精心标注,涵盖活动、网关、参与者及流程信息。创建过程中,首先对文本进行预处理和标注,随后通过自动化工具修正标注错误,并计算专家标注的一致性。PET数据集的应用领域主要集中在业务流程管理和信息提取,旨在通过提供高质量的标注数据,推动数据驱动的方法在业务流程提取中的应用,并促进不同方法之间的客观比较。
The PET Dataset was co-created by the Bruno Kessler Foundation and its partners, focusing on extracting business process models from natural language texts. This dataset contains 45 text descriptions, each meticulously annotated with information covering activities, gateways, participants and process details. During its development, texts were first preprocessed and annotated, followed by correction of annotation errors via automated tools and calculation of inter-annotator agreement among experts. The PET Dataset is mainly applied in the fields of business process management and information extraction. It aims to promote the application of data-driven approaches in business process extraction by providing high-quality annotated data, and facilitate objective comparisons between different methods.




