Performance evaluation and work-load estimation for geographic information systems

Abstract

Agencies acquiring GIS hardware and software are faced with uncertainty at two levels: over the degree to which the proposed system will perform the functions required, and over the degree to which it is capable of doing so within proposed production schedules. As the field matures the second concern is becoming more significant. A formal model of the process of acquiring a GIS is presented, based on the conceptual level of defining GIS sub-tasks. The appropriateness of the approach is illustrated using performance data from the Canada Land Data System. It is possible to construct reasonably accurate models of system resource utilization using simple predictors and least squares techniques, and a combination of inductive and deductive reasoning. The model has been implemented in an interactive package for MS-DOS systems.