Enhancing the Expressivity of Temporal Graph Networks through Source-Target Identification
This addresses a specific limitation in temporal graph modeling for dynamic node affinity prediction, representing an incremental improvement.
The paper tackled the problem of Temporal Graph Networks (TGNs) performing poorly on dynamic node affinity prediction, where simple heuristics like persistent forecasts and moving averages outperform them, and proposed TGNv2 by adding source-target identification to messages, which significantly outperforms TGN and other models on Temporal Graph Benchmark datasets.
Despite the successful application of Temporal Graph Networks (TGNs) for tasks such as dynamic node classification and link prediction, they still perform poorly on the task of dynamic node affinity prediction -- where the goal is to predict 'how much' two nodes will interact in the future. In fact, simple heuristic approaches such as persistent forecasts and moving averages over ground-truth labels significantly and consistently outperform TGNs. Building on this observation, we find that computing heuristics over messages is an equally competitive approach, outperforming TGN and all current temporal graph (TG) models on dynamic node affinity prediction. In this paper, we prove that no formulation of TGN can represent persistent forecasting or moving averages over messages, and propose to enhance the expressivity of TGNs by adding source-target identification to each interaction event message. We show that this modification is required to represent persistent forecasting, moving averages, and the broader class of autoregressive models over messages. Our proposed method, TGNv2, significantly outperforms TGN and all current TG models on all Temporal Graph Benchmark (TGB) dynamic node affinity prediction datasets.