Research On SQL-Natural Language Automatic Translation

SQL, namely Structured Query Language, is a special-purpose programming language designed for managing data held in a relational database management system(RDBMS). The power of SQL lies on its ability to link information from multiple tables to perform complex sets of procedures with a single statement.Unfortunately, while powerful, it is difficult to understand SQL for those persons with little or no experience with databases or computer science. However, in many circumstances, it is necessary to make SQL easily understood by non-computer professionals.so research on SQL to natural language translation has the very greatly practical significance.We hava not use complex natural language processing technology to implement the translation from SQL language to natural language because of lacking of SQL dataset. However, we found that SQL truly executive sequence is coherent to reading habits of human beings.Based on this; we presented an innovative method to implement the translation from SQL to natural language, which is accordance with SQL truly executive sequence.the main research work are as follows:(1) Design SQL lexer rules and SQL parser rules, then we make use of ANTLR to convert SQL sentence into abstract syntax tree.(2) Design corresponding simplified rules to make the final translation more human-readbale.(3) Develop practical translation system to validate effectiveness of our method.(4) Applied SQL analysis technology and SQL translation technology to database firewall and audit system to help the security of database.The experiments prove that our method could get a concise and understandable translation.