A METHODOLOGY TO ESTABLISH A UNIVERSAL DATA MODEL FOR GIS DATABASES

摘要

GIS is a multidisciplinary technology, with users spreading across various backgrounds. A database designed for a GIS would therefore, have to cater to the requirements of a wide range of technocrats. The feeder technologies to a GIS database such as surveying, satellite remote sensing, photogrammetry, computer science, CAD/CAM, etc., are also equally rich in variety. Added to all this, GISs have become enterprise-wide solutions across the Internet or Intranets. Further, it is very costly to change the design of existing databases, every time technology provides a major breakthrough. The process of GIS database design has thus become more complex that ever before. This presentation will provide a methodology to work out a robust and universal data model for designing a GIS database such that it would at least be - shareable, expandable and evolvable. The constructs of the model would exemplify with lessons learned from real life projects, how geographic data sets of the past centuries could be maintained, while at the same time, how the requirements of the present and future centuries (or should we say millennium) could be met. This model is not bound by variations in input data sources, application software, operating systems, computer hardware, digital data formats, multi-departmental mandates, users’ wish lists, etc. Such a data model would be able to accommodate historical paper maps of the 1800s, the digital images of today, intelligent object features of the near future, and hopefully any other innovative data set that comes up in the future. This presentation addresses technical, policy and organizational issues that may govern the development of a GIS. It is for an intermediate audience in terms of technology, but of great value to technocrats, who are the decisionmakers in their organization, especially for those who have to manage funding and get the biggest bang for the buck.