Graph is a Substrate Across Data Modalities
This addresses the issue of inefficiently reconstructing structural regularities for researchers and practitioners in graph representation learning, though it appears incremental as it builds on existing multi-task and representation learning concepts.
The paper tackles the problem of learning graph representations in an isolated manner across different data modalities and tasks, proposing G-Substrate as a framework to treat graph structure as a persistent substrate, which outperforms task-isolated and naive multi-task learning methods in experiments across multiple domains.
Graphs provide a natural representation of relational structure that arises across diverse domains. Despite this ubiquity, graph structure is typically learned in a modality- and task-isolated manner, where graph representations are constructed within individual task contexts and discarded thereafter. As a result, structural regularities across modalities and tasks are repeatedly reconstructed rather than accumulated at the level of intermediate graph representations. This motivates a representation-learning question: how should graph structure be organized so that it can persist and accumulate across heterogeneous modalities and tasks? We adopt a representation-centric perspective in which graph structure is treated as a structural substrate that persists across learning contexts. To instantiate this perspective, we propose G-Substrate, a graph substrate framework that organizes learning around shared graph structures. G-Substrate comprises two complementary mechanisms: a unified structural schema that ensures compatibility among graph representations across heterogeneous modalities and tasks, and an interleaved role-based training strategy that exposes the same graph structure to multiple functional roles during learning. Experiments across multiple domains, modalities, and tasks show that G-Substrate outperforms task-isolated and naive multi-task learning methods.