Developing a natural language interface to complex data
ACM Transactions on Database Systems (TODS)
| Hi-index | 0.00 |
This paper presents a method of formally representing the information that exists in a relational database. The primary utility of such a representation is for deductive question-answering systems that must access an existing relational database. To respond intelligently to user inquiries, such systems must have a more complete representation of the domain of discourse than is generally available in the database. The problem that then arises is how to reconcile the information present in the database with the domain representation, so that database queries can be derived to answer the user's inquiries. Here we take the formal approach of describing a relational database as the model of a first-order language. Another first-order language, the metalanguage, is used both to represent the domain of discourse, and to describe the relationship of the database to the domain. This view proves particularly useful in two respects. First, by axiomatiiing the database language and its associated model in a metatheory, we are able to describe in a powerful and flexible manner how the database corresponds to the domain of discourse. Secondly, viewing the database as a mechanizable model of the database language enables us to take advantage of the computational properties of database query language processors. Once a database query that is equivalent to an original query is derived, it can be evaluated against the database to determine the truth of the original query. Thus the algebraic operations of the database processor can be incorporated in an elegant way into the deductive process of question-answering.