FlowNet: Modeling Dynamic Spatio-Temporal Systems via Flow Propagation

2025-11-05

AI Summary

Existing dynamic spatiotemporal modeling approaches predominantly rely on similarity-driven graph connectivity or attention mechanisms, which struggle to capture asymmetric flow exchanges between nodes governed by physical conservation laws. To address this, we propose a novel “spatiotemporal flow” paradigm grounded in physics-informed flow conservation constraints. Our framework introduces flow tokens, a flow allocation module, and an adaptive spatial mask to enable context-aware, interpretable modeling of dynamic interactions. A cascaded architecture further enhances nonlinear representational capacity. Extensive experiments across three real-world systems and seven evaluation metrics demonstrate significant improvements over state-of-the-art methods—achieving higher modeling accuracy while preserving physical interpretability through explicit conservation-aware flow dynamics.

Technology Category

• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
• Computer Vision: Diffusion Models for Vision

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization

Abstract

Accurately modeling complex dynamic spatio-temporal systems requires capturing flow-mediated interdependencies and context-sensitive interaction dynamics. Existing methods, predominantly graph-based or attention-driven, rely on similarity-driven connectivity assumptions, neglecting asymmetric flow exchanges that govern system evolution. We propose Spatio-Temporal Flow, a physics-inspired paradigm that explicitly models dynamic node couplings through quantifiable flow transfers governed by conservation principles. Building on this, we design FlowNet, a novel architecture leveraging flow tokens as information carriers to simulate source-to-destination transfers via Flow Allocation Modules, ensuring state redistribution aligns with conservation laws. FlowNet dynamically adjusts the interaction radius through an Adaptive Spatial Masking module, suppressing irrelevant noise while enabling context-aware propagation. A cascaded architecture enhances scalability and nonlinear representation capacity. Experiments demonstrate that FlowNet significantly outperforms existing state-of-the-art approaches on seven metrics in the modeling of three real-world systems, validating its efficiency and physical interpretability. We establish a principled methodology for modeling complex systems through spatio-temporal flow interactions.

Problem

• Modeling dynamic spatio-temporal systems with flow-mediated interdependencies
• Capturing asymmetric flow exchanges governing system evolution
• Ensuring state redistribution aligns with physical conservation laws

Innovation

• Models dynamic couplings via quantifiable flow transfers
• Uses flow tokens as information carriers for transfers
• Dynamically adjusts interaction radius with adaptive masking