MyNLIDB: A Natural Language Interface to Database

Abstract

Structured Query Language (SQL) is a powerful tool to retrieve or manage data held in a relational database management system. But only the person who has a good knowledge of the SQL language can interact with the database. So to allow the users who don't know SQL, Natural language Interface to Database (NLIDB) systems are being developed to perform querying in natural language (NL) with the databases. While constructing an NLIDB, it is very critical to bridge the semantic gap between a natural language query (NLQ) and the underlying data. Keyword mapping, which is the task of mapping individual keywords in the original NLQ to database elements such as relations, attributes or values of attributes is a method through which this bridging is exhibited. As it is very difficult to understand the true meaning of every word in NLQ because of the ambiguity involved in NL, to map individual keywords to the schema definition and contents of the underlying database is a challenging task. In this paper, we present a system MyNLIDB that has a good performance with respect to keyword mapping. It uses a Schema-Graph created from the underlying database, Stanford part-of-speech parser and dependency parser to convert NL Query to SQL using pipeline processing. MyNLIDB is domain independent and database independent. It works very well for simple queries.