Full-History Graphs with Edge-Type Decoupled Networks for Temporal Reasoning

2025-08-05

AI Summary

To address the challenge of modeling temporal interactions among entities in dynamic graphs, this paper proposes the Full-History Graph—a novel representation that explicitly decouples static structural edges from inter-timestep self-connections, enabling joint spatiotemporal reasoning within a unified graph structure. We further design a dual-path attention mechanism: a spatial path employs graph attention to aggregate neighborhood relations, while a temporal path leverages multi-head temporal attention to capture evolutionary patterns; these are synergistically optimized via a fusion module. Evaluated on the Waymo driving intention prediction task, our method achieves a 75.6% joint accuracy; on the Elliptic++ Bitcoin fraud detection task, it attains an F1-score of 88.1%, significantly outperforming state-of-the-art baselines. The core contributions are (i) a spatiotemporal decoupling modeling paradigm and (ii) a parallel spatiotemporal message aggregation architecture.

Technology Category

• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning
• Planning, Routing, and Scheduling: Temporal Planning
• Data Mining & Knowledge Management: Mining of Spatial, Temporal or Spatio-Temporal Data

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
• Search and Retrieval-Augmented AI: Web query analysis, representation and understanding

Abstract

Modeling evolving interactions among entities is critical in many real-world tasks. For example, predicting driver maneuvers in traffic requires tracking how neighboring vehicles accelerate, brake, and change lanes relative to one another over consecutive frames. Likewise, detecting financial fraud hinges on following the flow of funds through successive transactions as they propagate through the network. Unlike classic time-series forecasting, these settings demand reasoning over who interacts with whom and when, calling for a temporal-graph representation that makes both the relations and their evolution explicit. Existing temporal-graph methods typically use snapshot graphs to encode temporal evolution. We introduce a full-history graph that instantiates one node for every entity at every time step and separates two edge sets: (i) intra-time-step edges that capture relations within a single frame and (ii) inter-time-step edges that connect an entity to itself at consecutive steps. To learn on this graph we design an Edge-Type Decoupled Network (ETDNet) with parallel modules: a graph-attention module aggregates information along intra-time-step edges, a multi-head temporal-attention module attends over an entity's inter-time-step history, and a fusion module combines the two messages after every layer. Evaluated on driver-intention prediction (Waymo) and Bitcoin fraud detection (Elliptic++), ETDNet consistently surpasses strong baselines, lifting Waymo joint accuracy to 75.6% (vs. 74.1%) and raising Elliptic++ illicit-class F1 to 88.1% (vs. 60.4%). These gains demonstrate the benefit of representing structural and temporal relations as distinct edges in a single graph.

Problem

• Modeling evolving entity interactions in temporal graphs
• Decoupling structural and temporal relations for accurate reasoning
• Improving prediction in driver maneuvers and financial fraud detection

Innovation

• Full-history graph with separate edge sets
• Edge-Type Decoupled Network (ETDNet) design
• Parallel modules for intra and inter-time-step edges