Disentangled Instrumental Variables for Causal Inference with Networked Observational Data

2026-02-08

AI Summary

This work addresses the challenge posed by unobserved confounding in network observational data, which often violates the exogeneity assumption of instrumental variables (IVs). Existing methods struggle to disentangle environment-shared endogenous correlations from individual-specific exogenous variation. To overcome this limitation, we propose DisIV, a novel framework that, for the first time, enables decoupled extraction of valid IVs from network data with latent confounders. By leveraging network homophily as an inductive bias, DisIV introduces a structural decoupling mechanism to isolate individual-specific components as candidate IVs, while enforcing orthogonality and exclusion constraints to ensure causal validity. Extensive semi-synthetic experiments on multiple real-world datasets demonstrate that DisIV significantly outperforms state-of-the-art baselines, effectively mitigating confounding dependencies induced by neighbor-based modeling.

Technology Category

• Constraint Satisfaction and Optimization: Distributed CSP/Optimization
• Reasoning under Uncertainty: Causality
• Machine Learning: Causal Learning

Application Category

• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Web Mining and Content Analysis: Bridging structured and unstructured data
• Economics, Online Markets and Human Computation: Incentives in network design for Web infrastructures and ecosystems

Abstract

Instrumental variables (IVs) are crucial for addressing unobservable confounders, yet their stringent exogeneity assumptions pose significant challenges in networked data. Existing methods typically rely on modelling neighbour information when recovering IVs, thereby inevitably mixing shared environment-induced endogenous correlations and individual-specific exogenous variation, leading the resulting IVs to inherit dependence on unobserved confounders and to violate exogeneity. To overcome this challenge, we propose $\underline{Dis}$entangled $\underline{I}$nstrumental $\underline{V}$ariables (DisIV) framework, a novel method for causal inference based on networked observational data with latent confounders. DisIV exploits network homogeneity as an inductive bias and employs a structural disentanglement mechanism to extract individual-specific components that serve as latent IVs. The causal validity of the extracted IVs is constrained through explicit orthogonality and exclusion conditions. Extensive semi-synthetic experiments on real-world datasets demonstrate that DisIV consistently outperforms state-of-the-art baselines in causal effect estimation under network-induced confounding.

Problem

• Instrumental Variables
• Networked Observational Data
• Unobserved Confounders
• Exogeneity
• Causal Inference

Innovation

• Disentangled Instrumental Variables
• Networked Observational Data
• Latent Confounders
• Structural Disentanglement
• Causal Inference