Hierarchical Graph Neural Networks for Causal Discovery and Root Cause Localization

TL;DR

This paper introduces REASON, a hierarchical graph neural network framework that automatically discovers intra- and inter-network causal relationships for root cause localization in complex systems.

Contribution

The paper presents a novel hierarchical GNN-based framework combining topological and individual causal discovery for effective root cause analysis.

Findings

Effective in localizing root causes in real-world datasets

Outperforms existing methods in accuracy and robustness

Demonstrates scalability to large systems

Abstract

In this paper, we propose REASON, a novel framework that enables the automatic discovery of both intra-level (i.e., within-network) and inter-level (i.e., across-network) causal relationships for root cause localization. REASON consists of Topological Causal Discovery and Individual Causal Discovery. The Topological Causal Discovery component aims to model the fault propagation in order to trace back to the root causes. To achieve this, we propose novel hierarchical graph neural networks to construct interdependent causal networks by modeling both intra-level and inter-level non-linear causal relations. Based on the learned interdependent causal networks, we then leverage random walks with restarts to model the network propagation of a system fault. The Individual Causal Discovery component focuses on capturing abrupt change patterns of a single system entity. This component examines the temporal patterns of each entity's metric data (i.e., time series), and estimates its likelihood of being a root cause based on the Extreme Value theory. Combining the topological and individual causal scores, the top K system entities are identified as root causes. Extensive experiments on three real-world datasets with case studies demonstrate the effectiveness and superiority of the proposed framework.