Enriching the class diagram concepts to capture natural language semantics for database access

Abstract

Research on accessing databases using natural languages usually utilizes an intermediate logical form for the mapping process from natural languages to database query languages. However, there is still much that needs to be accomplished to bridge the gap between natural language constructs and database schemas. In this paper, we present a translation scheme for transforming natural language queries into relational algebra through the class diagram representations. Based on a logical form developed by extending the UML class diagram notations, a transformation model is presented to support the automatic transformation of natural language queries into relational algebra by employing appropriate natural language processing techniques and object-oriented analysis methods. The proposed logical form has the advantage that it can be mapped from natural language constructs by referring to the conceptual schema modeled by class diagrams, and can be efficiently transformed into relational algebra for query execution. We believe the whole process could offer a clear and natural framework for processing natural language queries to retrieve data from database systems.