Integrating Background Knowledge for Scalable Causal Discovery

2026-07-11

AI Summary

This work addresses the challenges of high computational complexity and structural identifiability in large-scale causal discovery. It proposes a novel framework that dynamically integrates expert background knowledge into the core search process of scalable causal discovery—rather than merely using it for post-processing—for the first time. By combining constraint-based reasoning with efficient graph learning algorithms, the method actively guides structural exploration during the search, substantially reducing the hypothesis space. Experimental results demonstrate that the framework not only lowers computational overhead but also significantly improves the accuracy and identifiability of the inferred causal graphs, making it particularly well-suited for large-scale scenarios where only partial causal structures need to be recovered.

Abstract

Expert background knowledge is often available in practical applications of causal discovery. Such constraints on the true causal graph can help causal discovery in terms of identifiability of causal effects and accuracy of the learned structure, but also in reducing the space of candidate causal graphs. As causal discovery can become computationally expensive for large number of variables, it is crucial to utilize background knowledge effectively during the causal discovery process. However, most current methods only use background knowledge in a postprocessing step after causal discovery to refine the learned graph. In this work, we develop a framework for utilizing background knowledge during the causal discovery process, focusing especially on scalable causal discovery methods that recover only a subset of the whole graph. We implement our framework for multiple algorithms and empirically show that utilizing background knowledge can both reduce computational requirements and increase the quality of the learned structures.

Problem

• causal discovery
• background knowledge
• scalability
• computational efficiency
• causal graph

Innovation

• causal discovery
• background knowledge
• scalable algorithms
• graph constraints
• computational efficiency