TrendGNN: Towards Understanding of Epidemics, Beliefs, and Behaviors

2025-11-29

AI Summary

This study addresses the interpretability challenge in forecasting human beliefs and behavioral signals during epidemics. Methodologically, it constructs a trend-similarity graph capturing dynamic temporal correlations among multi-source signals—including epidemiological data, public beliefs, and real-world behaviors—and designs an interpretable graph neural network (GNN) to automatically identify key predictive signals and their cross-modal contribution pathways. Its core innovation lies in integrating time-varying trend similarity modeling into graph structure learning, enabling explicit disentanglement and attribution of nonlinear, time-dependent inter-signal dependencies. Experiments demonstrate that the framework significantly improves both prediction accuracy and interpretability for belief and behavior signals, enhances the reliability of intervention impact estimation, and provides interpretable, grounded representations for integrative epidemic simulation modeling. (149 words)

Technology Category

• Knowledge Representation and Reasoning: Reasoning with Beliefs
• Reasoning under Uncertainty: Graphical Models
• Machine Learning: Graph-based Machine Learning

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization

Abstract

Epidemic outcomes have a complex interplay with human behavior and beliefs. Most of the forecasting literature has focused on the task of predicting epidemic signals using simple mechanistic models or black-box models, such as deep transformers, that ingest all available signals without offering interpretability. However, to better understand the mechanisms and predict the impact of interventions, we need the ability to forecast signals associated with beliefs and behaviors in an interpretable manner. In this work, we propose a graph-based forecasting framework that first constructs a graph of interrelated signals based on trend similarity, and then applies graph neural networks (GNNs) for prediction. This approach enables interpretable analysis by revealing which signals are more predictable and which relationships contribute most to forecasting accuracy. We believe our method provides early steps towards a framework for interpretable modeling in domains with multiple potentially interdependent signals, with implications for building future simulation models that integrate behavior, beliefs, and observations.

Problem

• Forecasting epidemic signals with interpretability
• Modeling interplay between epidemics, beliefs, behaviors
• Using graph neural networks for interpretable predictions

Innovation

• Graph-based forecasting framework using trend similarity
• Graph neural networks for interpretable epidemic prediction
• Reveals key signals and relationships for forecasting accuracy