Probabilities of Causation and Root Cause Analysis with Quasi-Markovian Models

2025-09-02

AI Summary

To address theoretical non-identifiability and high computational complexity in causal probability estimation—arising from partial identifiability and latent confounding—this paper proposes an efficient, systematic root-cause analysis framework. Methodologically, it integrates quasi-Markov models, structural constraints from causal graphs, probabilistic bounding inference, and latent-variable elimination techniques to design a lightweight algorithm that significantly reduces the computational overhead of causal effect bound estimation. A key innovation is the first-ever interpretable ranking of *entire causal paths*—rather than individual edges or nodes—enabling precise attribution of multi-path causal contributions. Experiments demonstrate that, while preserving theoretical rigor, the framework improves bound computation efficiency by one to two orders of magnitude and achieves robust, scalable causal tracing in complex systems.

Technology Category

• Reasoning under Uncertainty: Causality
• Machine Learning: Causal Learning
• Knowledge Representation and Reasoning: Action, Change, and Causality

Application Category

• Graph Algorithms and Modeling for the Web: Algorithms and analysis for incomplete, noisy, or partially observed Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web

Abstract

Probabilities of causation provide principled ways to assess causal relationships but face computational challenges due to partial identifiability and latent confounding. This paper introduces both algorithmic simplifications, significantly reducing the computational complexity of calculating tighter bounds for these probabilities, and a novel methodological framework for Root Cause Analysis that systematically employs these causal metrics to rank entire causal paths.

Problem

• Reducing computational complexity of causal probability bounds
• Addressing partial identifiability and latent confounding challenges
• Developing systematic root cause analysis using causal metrics

Innovation

• Algorithmic simplifications reduce computational complexity
• Novel framework ranks causal paths systematically
• Quasi-Markovian models address latent confounding