Uncovering Causal Relation Shifts in Event Sequences under Out-of-Domain Interventions

2025-07-14

AI Summary

In real-world scenarios, cross-domain exogenous interventions dynamically alter the causal structure of event sequences, yet existing methods typically assume i.i.d. data and neglect such effects. Method: We propose the first time-series causal inference framework explicitly designed for out-of-domain interventions. It introduces and estimates the temporal Average Treatment Effect (ATE), extending the Rubin causal model beyond stationary causal assumptions; employs a Transformer-based dual-path neural architecture to jointly model long-range dependencies, local patterns, and intervention signals; and constructs an unbiased ATE estimator to ensure identifiability and quantification of causal shifts. Results: Extensive experiments across healthcare, manufacturing, and transportation domains demonstrate significant improvements in ATE estimation accuracy and point-process modeling performance over state-of-the-art baselines.

Technology Category

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

Application Category

• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web

Abstract

Inferring causal relationships between event pairs in a temporal sequence is applicable in many domains such as healthcare, manufacturing, and transportation. Most existing work on causal inference primarily focuses on event types within the designated domain, without considering the impact of exogenous out-of-domain interventions. In real-world settings, these out-of-domain interventions can significantly alter causal dynamics. To address this gap, we propose a new causal framework to define average treatment effect (ATE), beyond independent and identically distributed (i.i.d.) data in classic Rubin's causal framework, to capture the causal relation shift between events of temporal process under out-of-domain intervention. We design an unbiased ATE estimator, and devise a Transformer-based neural network model to handle both long-range temporal dependencies and local patterns while integrating out-of-domain intervention information into process modeling. Extensive experiments on both simulated and real-world datasets demonstrate that our method outperforms baselines in ATE estimation and goodness-of-fit under out-of-domain-augmented point processes.

Problem

• Identifying causal shifts in event sequences under external interventions
• Measuring treatment effects beyond traditional i.i.d. data assumptions
• Modeling temporal dependencies with out-of-domain intervention impacts

Innovation

• Proposes new causal framework for out-of-domain interventions
• Designs unbiased ATE estimator for temporal processes
• Uses Transformer model to integrate intervention information