AIOps大型真实世界基准数据集
收藏资源简介:
本研究介绍了三个大型真实世界数据集,旨在支持AIOps领域的关键任务,包括KPI异常检测、多维数据根因定位及故障发现与诊断。这些数据集由清华大学等机构与工业合作伙伴共同创建,涵盖了多个互联网公司的生产系统数据。通过这些数据集,研究者们能够更有效地开发和评估AIOps技术,同时,这些数据集也支持了年度AIOps算法挑战赛,促进了学术界与工业界的合作与交流。
This study introduces three large-scale real-world datasets developed to support core tasks in the field of AIOps, including KPI anomaly detection, multi-dimensional data root cause localization, and fault discovery and diagnosis. Jointly created by institutions including Tsinghua University and industrial partners, these datasets cover production system data from multiple Internet companies. These datasets enable researchers to develop and evaluate AIOps technologies more efficiently; furthermore, they have supported the annual AIOps Algorithm Challenge, fostering cooperation and exchanges between academia and industry.



