Intelligent Database Query Translation

In this paper we are concerned with techniques for processing database queries that are not required to be formulated based on complete and precise knowledge of the query specification as defined in the database management system. In specific, we consider processing queries that are ill-defined, not complete and fuzzy . We propose a query translator that tries to bridge the gap between the user's semantically stated query and its full specification as required by the database. The translator utilizes a graph based structure to represent queries and relations in the database. The translation process involves transforming incomplete query graphs into complete query graphs using attribute covering where the entities of the query are mapped to real entities of the database and functional dependency resolution, where the relational joins between these entities are established. The proposed translator is an attempt at incorporating intelligence in query processing systems to support user- friendly features such as cooperative and corrective responses from database systems.