A Relational Approach to Vector Data Structure Conversion

The proliferation of geographic information systems and digital data bases is creating a need for efficient meth ods to convert data from one spatial data structure to another. One approach is to create ad hoc interfaces, with a potential of N(N-1) interfaces for N data struc tures. Using an intermediate data structure, at most 2N interfaces are required. An intermediate relational data structure is therefore proposed that takes the form of a set of normalized relations stored in a relational infor mation management system. The advantages of this approach are found in the simplicity of the relational approach, and the availability of relational operators to be used as higher level tools, to convert from and to the relational data structure. The Relational Information Management System (RIM) is used for the ongoing research. In con junction with the relational data structure, another higher level tool has been developed to cope with linked lists, tree structures, vectors, and matrices, which are not otherwise easily reduced. This data tracking system is programmable at a higher level, in a syntax that allows a concise expression of the desired restructuring. Output from this system can be further operated on by relational operators to arrive at the desired intermediate data structure. This, and other topology verification and checking tools, are visualized as part of a core system dedicated to the conversion and collation of spatial data from diverse origins. *Work performed under U.S. Geological Survey contract no. 14-08-0001-20129.