In view of the great bulk of information available on the World Wide Web, performing searches and retrieving relevant information now constitute a tremendous challenge. Current search engines have popularized the use of natural language words in queries. Relational databases are employed on a large-scale basis to store information on a wide range of subjects, but retrieving such information through the Web faces several difficulties, related to the need for schema knowledge and access language and, above all, the need to produce a coherent interpretation of user query for semantically equivalent queries to be run in the database. This paper aims to discuss the development and main results of an approach that interprets natural language queries and the use of domain ontologies as a tool that transforms a formulated query into a database access language.