Abstract Event Causal Rules: Induction and Application

2026-08-05

AI Summary

This work addresses the limited generalization of existing event causality models on low-frequency, long-tail, and unseen event combinations, which hinders their applicability to tasks such as risk预警. The authors propose the Abstract Event Causal Rule (AECR) paradigm, introducing for the first time a relation-level abstract causal representation. By leveraging multi-agent causal induction and similarity-constrained clustering, the method distills noise-robust, generalizable causal logic from raw data and constructs the first AECR knowledge base. Furthermore, a rule-guided causal attention encoder is designed, integrating gated representation fusion with a graph-based event prediction model to inject AECR knowledge into downstream tasks. Experiments demonstrate that the proposed approach substantially enhances the generalization capability of event causal reasoning, achieving consistent and significant performance gains in predicting rare and previously unseen events.

Abstract

Event-centric intelligent analytical systems heavily depend on explicit causal event knowledge for risk early warning, decision-making support and narrative comprehension. Nevertheless, existing instance-level causal pairs suffer severe generalization deficits on low-frequency long-tail and unseen event combinations. To address this limitation, this work proposes Abstract Event Causal Rule (AECR), a novel relation-level causal abstraction paradigm that transforms concrete cause-effect pairs into generalized abstract causal logic while retaining their intrinsic causal relationships. We design a multi-agent Concrete-to-Abstract Causal Induction (CACI) system coupled with similarity-constrained clustering to distill trustworthy AECRs from noisy raw causal data, based on which two complete AECR knowledge bases are built. To validate the practical utility of abstract causal knowledge, we propose an Abstract Rule-Guided Causal Attention Encoder (AR-GCAE), which injects the retrieved AECRs into the causality Graph Event Prediction (CGEP) benchmark task via rule-guided attention layers and gated representation fusion. Quantitative experimental results reveal that applying AECRs substantially strengthens the generalization capacity of event causal reasoning and brings consistent performance improvements to event prediction, with the most prominent gains observed on rare and unseen event samples.

Problem

• event causality
• generalization deficit
• long-tail events
• unseen event combinations
• causal knowledge

Innovation

• Abstract Event Causal Rule
• Causal Abstraction
• Multi-agent Induction
• Rule-guided Attention
• Event Prediction