REDELEX: A Framework for Relational Deep Learning Exploration
This work addresses a gap in understanding performance drivers for relational deep learning, which is incremental as it builds on the existing RDL paradigm by providing systematic evaluation insights.
The authors tackled the lack of analysis linking Relational Deep Learning (RDL) model performance to relational database (RDB) characteristics by introducing REDELEX, a framework that evaluates RDL models on over 70 RDBs, confirming RDL's generally superior performance over classic methods and identifying key factors like model complexity and database properties.
Relational databases (RDBs) are widely regarded as the gold standard for storing structured information. Consequently, predictive tasks leveraging this data format hold significant application promise. Recently, Relational Deep Learning (RDL) has emerged as a novel paradigm wherein RDBs are conceptualized as graph structures, enabling the application of various graph neural architectures to effectively address these tasks. However, given its novelty, there is a lack of analysis into the relationships between the performance of various RDL models and the characteristics of the underlying RDBs. In this study, we present REDELEX$-$a comprehensive exploration framework for evaluating RDL models of varying complexity on the most diverse collection of over 70 RDBs, which we make available to the community. Benchmarked alongside key representatives of classic methods, we confirm the generally superior performance of RDL while providing insights into the main factors shaping performance, including model complexity, database sizes and their structural properties.