RealTCD: Temporal Causal Discovery from Interventional Data with Large Language Model
Peiwen Li SIGS, Tsinghua University , Xin Wang DCST, BNRist, Tsinghua University , Zeyang Zhang DCST, Tsinghua University , Yuan Meng DCST, Tsinghua University , Fang Shen Alibaba Cloud , Yue Li Alibaba Cloud , Jialong Wang Alibaba Cloud , Yang Li SIGS, Tsinghua University and Wenwu Zhu DCST, BNRist, Tsinghua University
Abstract
In the field of Artificial Intelligence for Information Technology Operations, causal discovery is pivotal for operation and maintenance of systems, facilitating downstream industrial tasks such as root cause analysis. Temporal causal discovery, as an emerging method, aims to identify temporal causal relations between variables directly from observations by utilizing interventional data. However, existing methods mainly focus on synthetic datasets with heavy reliance on interventional targets and ignore the textual information hidden in real-world systems, failing to conduct causal discovery for real industrial scenarios. To tackle this problem, in this paper we investigate temporal causal discovery in industrial scenarios, which faces two critical challenges: how to discover causal relations without the interventional targets that are costly to obtain in practice, and how to discover causal relations via leveraging the textual information in systems which can be complex yet abundant in industrial contexts. To address these challenges, we propose the RealTCD framework, which is able to leverage domain knowledge to discover temporal causal relations without interventional targets. W
中文速览
在复杂的工业运维系统（AIOps）中，准确找出指标之间的时序因果关系（temporal causal discovery）是做根因分析的基础，但现有方法依赖昂贵且难以获取的"干预目标"标签，也完全忽视了系统中大量存在的文本信息。为此，作者提出了 RealTCD 框架，核心包含两个模块：一是基于打分的时序因果发现算法，通过对邻接矩阵和干预族同时施加正则化并联合优化，在不需要人工标注干预目标的情况下仍能识别真实因果图；二是 LLM 引导的元初始化模块，利用大语言模型解析系统中的文本描述，将领域知识转化为因果结构的初始化先验，从而提升发现质量。在模拟数据集和真实工业场景的双重实验中，RealTCD 均显著优于现有基线方法，证明了该框架在真实工业环境中落地的可行性和有效性。
原文 arXiv:2404.14786;中英对照 + 大白话阅读 https://aha.fim.ai/paper/2404.14786v2