Relational Databases
Overview
Relational databases organize information into tables connected by keys, with SQL and integrity rules supporting reliable storage, updates, and analysis. Research improves how systems plan and tune queries, protect sensitive data, and integrate or document schemas; newer work also uses language models and agents to translate questions into SQL and explore databases. For learning from relational data, approaches include graph neural networks, relational transformers, and synthetic pretraining, evaluated on tasks such as RelBench as well as SQL benchmarks like Spider and BIRD. A central challenge is making systems dependable across large, unfamiliar schemas and SQL dialects while balancing accuracy, execution cost, privacy, and the complexity of specialized methods.
Latest papers
80 papers5 papers in the last four weeks, down 50% on the four weeks before. 0.0% of all new papers.