Is Fixing Schema Graphs Necessary? Full-Resolution Graph Structure Learning for Relational Deep Learning

Relational deep learning models, which apply graph neural networks to relational databases, often rely on a design principle that prioritizes full-resolution graph structure. However, this approach may not be strictly necessary, as recent research suggests that fixing schema graphs can be bypassed in favor of full-resolution graph structure learning. This alternative method allows for more flexible and adaptive modeling of complex relational data. By learning the graph structure directly from the data, rather than relying on predefined schema, models can better capture nuanced relationships and improve predictive performance. The implications of this approach are significant, as it can enhance the accuracy and effectiveness of relational deep learning models in a wide range of applications, from database querying to knowledge graph embedding¹. This development matters to practitioners because it can lead to more efficient and accurate relational deep learning models, with potential applications in areas such as data integration and database management.