Query relational databases in Punjabi language

Abstract

Public relational databases are accessed by end users to get the information they require. Direct interaction with relational databases requires the knowledge of structured query language (SQL). It is not feasible for every user to learn SQL. An access through an application limits the query options. An end user can ask a query very easily in a natural language. To provide the full advantages of public access, the users should be allowed to query the required data through natural language questions. It is possible by providing natural language support to query relational databases. This paper presents the system model, design and implementation to query relational databases in Punjabi language. It allows human–machine interaction in Punjabi language for information retrieval. It accepts a Punjabi language query in flexible format, uses pattern matching techniques to prepare an SQL query from it, maps data element tokens of the query to actual database objects and joins multiple tables to fetch the required data.