Text-to-SQL Generation

Text-to-SQL Generation

This project develops Generative AI models to translate natural language questions or commands into structured SQL queries. By bridging the gap between human language and database querying, this technology enables users without extensive SQL knowledge to easily retrieve information from databases. Our models interpret user intent, understand database schema, and generate accurate SQL queries, revolutionizing database interactions across various disciplines.

Director

  • Jinho Choi - Associate Professor at Emory University

Publication

  1. ESM+: Modern Insights into Perspective on Text-to-SQL Evaluation in the Age of Large Language Models. Ascoli, B.; Kandikonda, R.; Choi, J. D. arXiv, 2024.
  2. SMAT: An Attention-based Deep Learning Solution to the Automation of Schema Matching. Zhang, J.; Shin, B.; Choi, J. D.; and Ho, J. Proceedings of the European Conference on Advances in Databases and Information Systems (ADBIS), 2021.