Improving the customization of natural language interface to databases using an ontology

Natural language interfaces to databases are considered one of the best alternatives for final users who wish to make complex, uncommon and frequent queries, which is a very common need in organizations. The use of this type of interfaces has been very limited, due to their limited publicizing and the complexity to adapt them to users' needs, and because their precision varies widely. We propose as a solution to the problem of customizing this type of interfaces, the use of an ontology as a knowledge base whose design is simple and flexible enough to make the use and acceptance of these interfaces more accessible. This paper describes the design of the ontology, as well as a series of comparative evaluations of this approach versus the customization process of a commercial interface. This evaluation aims at assessing the acceptance of this approach by of those that will potentially customize the interface to a database, in contrast to the precision tests that are commonly applied to this type of interfaces. In spite of the difficulties found to carry out the evaluations, the results show that the use of our approach is preferred as a natural language interface customization means to the process of the most popular commercial interface. The estimations indicate that the potential people on charge of the process of customization of this type of interfaces considers that using the ontology as interface knowledge base would allow to answer a wider diversity of types of queries than those that would allow to answer a commercial interface.