170K Incident Reports
收藏arXiv2025-09-30 收录
下载链接：
https://github.com/RemilYoucef/split-sd4x
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含了过去7年中超过1000台服务器涉及的17万个事件报告，旨在评估子组模型在事件分拣中解释黑盒模型预测的有效性。此外，该数据集用于评估SplitSD4X方法，并将其与其他基线模型在事件分拣任务中进行比较。该数据集的规模达到17万个事件，任务重点在于事件分拣以及黑盒预测的解释。
This dataset contains 170,000 incident reports involving over 1,000 servers over the past seven years, aiming to evaluate the effectiveness of subgroup models in explaining black-box model predictions for incident triage tasks. Additionally, this dataset is utilized to assess the SplitSD4X method and compare it with other baseline models on the incident triage task. The dataset totals 170,000 incident entries, with the core tasks focusing on incident triage and the explanation of black-box model predictions.
