Geo-Relational Algebra: A Model and Query Language for Geometric Database Systems

Ralf Hartmut Güting

Fachbereich Informatik, Universität Dortmund, D-4600 Dortmund 50, West Germany

Abstract: The user's conceptual model of a database system for geometric data should be

simple and precise: easy to learn and understand, with clearly defined semantics,
expressive: allow to express with ease all desired query and data manipulation tasks,
efficiently implementable.

To achieve these goals we propose to extend relational database management systems by integrating geometry at all levels: At the conceptual level, relational algebra is extended to include geometric data types and operators. At the implementation level, the wealth of algorithms and data structures for geometric problems developed in the past decade in the field of Computational Geometry is exploited. - The paper starts from a view of relational algebra as a many-sorted algebra which allows to easily embed geometric data types and operators. A concrete algebra for two-dimensional applications is developed. It can be used as a highly expressive retrieval and data manipulation language for geometric as well as standard data. Finally, geo-relational database systems and their implementation strategy are discussed.

Published: In: J.W. Schmidt, S. Ceri, and M. Missikoff (eds.), Advances in Database Technology - EDBT '88. Proc. of the Intl. Conf. on Extending Database Technology, Venice, March 1988, 506-527.