MSAIL Natural Language to Code Project

This is a brief guide for the NL to Code project in MSAIL 2020-2021.

Getting Started

Here are a few resources for getting started with this project:

• Text2SQL Introduction Series - This is set of Medium posts that give a good introduction to the problem and approaches to turning natural language into SQL
• A Visual Guide to Using BERT for the First Time - A blog post introducing how to use BERT, the language model developed by Google
• BERT Github Repository
• BERT with Cloud TPU Colab - A Colab notebook to fine-tune and play around with BERT

Resources

Datasets:

• CoSQL dataset
• Sparc dataset
• Spider dataset

Some papers:

• Data-Anonymous Encoding for Text-to-SQL Generation
• NL2pSQL: Generating Pseudo-SQL Queries from Under-Specified Natural Language Questions
• CoSQL: A Conversational Text-to-SQL Challenge Towards Cross-Domain Natural Language Interfaces to Databases
• Machine Translation from Natural Language to Code using Long-Short Term Memory
• Mapping Language to Code in Programmatic Context