Bibliography on data base structures
ACM SIGMIS Database
| Hi-index | 0.00 |
This paper describes the file structures and access methods being used in the Geographic Data System of the Natural Information Systems Project at the University of British Columbia. These structures are designed to support a general input-output framework which is the foundation of each of the primary functions of the system. The paper outlines the basic functions of such a system and also the characteristics of geographic data which are unique. This produces a list of requirements for file structures and access methods. The paper describes the file structures and access methods chosen and also explains the ways in which the requirements are satisfied. The rules for the system are that it be easy to use, easy to expand and evolve, and interactive. This paper should be of interest because the structures and problems themselves are interesting, but also because more general principles of data manipulation can be extracted from the experience here. The majority of commercial installations are now considering interactive information manipulation of one form or another because of the rapidly changing equipment technology. In order to be useful, natural information systems appear to demand higher sophistication in their information manipulation than the standard commercial installation at its current stage of development. However the sophistication of the average shop is increasing rapidly, and the principles described in this paper are probably representative of problems which will be faced by many organizations in the near future. These broader considerations are outlined at the end of the paper.