A large crowd-sourced dataset for developing natural language interfaces for relational databases.