Generic querying of relational databases using natural language generation techniques

Abstract

This paper presents a method of querying databases by means of a natural languagelike interface which offers the advantage of minimal configuration necessary for porting the system. The method allows us to first automatically infer the set of possible queries that can apply to a given database, automatically generate a lexicon and grammar rules for expressing these queries, and then provide users with an interface that allows them to pose these queries in natural language without the well-known limitations of most natural language interfaces to databases. The way the queries are inferred and constructed means that semantic translation is performed with perfect reliability. © 2006 Association for Computational Linguistics.