Real-world Telecom Alarm Dataset

该数据集来自中国一个中等规模城市的主要移动运营商，包含了各种类型的警报记录和设备信息。其中，针对15种特定警报类型及其有向边，提供了真实的原因关系数据。此外，数据集还为6000个警报事务样本提供了真实的根本原因警报。该数据集的规模包括来自3818个设备的67万2639条警报记录，涵盖78种不同的警报类型。研究任务旨在进行根本原因警报定位和因果图学习。

This dataset is collected from the primary mobile operator of a medium-sized city in China, containing various types of alarm records and device information. Specifically, it provides real causal relationship data for 15 specific alarm types and their directed edges. Additionally, the dataset offers real root-cause alarms for 6,000 alarm transaction samples. In terms of scale, the dataset includes 672,639 alarm records from 3,818 devices, covering 78 distinct alarm types. The research tasks targeting this dataset include root-cause alarm localization and causal graph learning.

搜集汇总

背景与挑战

背景概述

该数据集名为'Real-world Telecom Alarm Dataset'，用于电信网络中的根因告警发现研究，与论文《An Influence-based Approach for Root Cause Alarm Discovery in Telecom Network》相关。