HGCN(O): A Self-Tuning GCN HyperModel Toolkit for Outcome Prediction in Event-Sequence Data

2025-07-30

AI Summary

This paper addresses two key challenges in event sequence prediction: insufficient graph-structure modeling and performance degradation due to class imbalance. To this end, we propose an Adaptive Graph Convolutional Network (GCN) toolkit. Our method unifies the modeling of heterogeneous graph structures, node- and graph-level attributes, and temporal dependencies. It supports dynamic selection among four GCN architectures—O-GCN, T-GCN, TP-GCN, and TE-GCN—and explicitly encodes temporal dynamics via edge-weight reconstruction using GCNConv and GraphConv layers. The framework demonstrates strong robustness on both balanced and imbalanced datasets: GCNConv significantly outperforms baselines under class imbalance, while all variants maintain stable, efficient performance on balanced data. Extensive experiments show that our approach consistently surpasses conventional time-series and graph-based models in tasks such as business process state prediction.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Reasoning under Uncertainty: Graphical Models
• Data Mining & Knowledge Management: Graph Mining, Social Network Analysis & Community

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

We propose HGCN(O), a self-tuning toolkit using Graph Convolutional Network (GCN) models for event sequence prediction. Featuring four GCN architectures (O-GCN, T-GCN, TP-GCN, TE-GCN) across the GCNConv and GraphConv layers, our toolkit integrates multiple graph representations of event sequences with different choices of node- and graph-level attributes and in temporal dependencies via edge weights, optimising prediction accuracy and stability for balanced and unbalanced datasets. Extensive experiments show that GCNConv models excel on unbalanced data, while all models perform consistently on balanced data. Experiments also confirm the superior performance of HGCN(O) over traditional approaches. Applications include Predictive Business Process Monitoring (PBPM), which predicts future events or states of a business process based on event logs.

Problem

• Predicts event sequences using self-tuning GCN models
• Optimizes accuracy for balanced and unbalanced datasets
• Outperforms traditional methods in outcome prediction

Innovation

• Self-tuning GCN toolkit for event prediction
• Integrates multiple graph representations and attributes
• Optimizes accuracy for balanced and unbalanced datasets