Relational Keyword Search (R-KwS) systems enable naive/informal users to retrieve information from relational databases without any knowledge about schema details or query languages. Roughly, R-KwS systems must determine which pieces of information to retrieve from the database and how to connect them to generate a relevant answer for the user. For instance, the keywords from the query may refer either to instance values or to database schema elements, such as relations and attributes names. Unfortunately, most of the R-KwS systems do not support references to the database schema. In this work, we propose a new method to generate an adequate SQL query from an input keyword query, taking advantage of references to the content of the database as well as references to the database schema. Our experiments indicate that schema references can help to improve the quality of the ranking the answers for the user.

Relational Keyword Search (R-KwS) systems enable naive/informal users to retrieve information from relational databases without any knowledge about schema details or query languages. Roughly, R-KwS systems must determine which pieces of information to retrieve from the database and how to connect them to generate a relevant answer for the user. For instance, the keywords from the query may refer either to instance values or to database schema elements, such as relations and attributes names. Unfortunately, most of the R-KwS systems do not support references to the database schema. In this work, we propose a new method to generate an adequate SQL query from an input keyword query, taking advantage of references to the content of the database as well as references to the database schema. Our experiments indicate that schema references can help to improve the quality of the ranking the answers for the user.