Graduate students Yuepeng Wang and Navid Yaghmazadeh, Associate Professor Işil Dillig and Assistant Professor Thomas Dillig won the best paper award at the 2017 Object-Oriented Programming, Systems, Languages & Applications (OOPSLA) conference for their paper “SQLizer: Query Synthesis from Natural Language.”
Their paper introduces a new technique, called SQLizer, for automatically turning natural language into SQL queries that is fully automated and works on any database. They tested SQLizer on three different databases: Microsoft Academic Search, Yelp and the Internet Movie Database. SQLizer ranked the solution to the desired query within the top five results 90 percent of the time, and the solution was ranked first almost 80 percent of the time. They also compared SQLizer to another natural language interface for databases, NaLIR, which won a best paper award at the Very Large Data Bases conference in 2014, and SQLizer outperformed NaLIR when tested on all three databases.