Year: 1983
Type: article
Abstract: Structuring database records by considering data item usage can yield substantial efficiencies in the operating cost of database systems.However, since the number of possible physical record structures for database of practical significance is enormous, and their evaluation is extremely complex, determining efficient record structures by full enumeration is generally infeasible.This paper discusses the techniques of mathematical clustering, iterative grouping refinement, mathematical programming, and hierarchic aggregation, which can be used to quickly determine efficient record structures for large, shared databases.
Source: ACM Computing Surveys
Author: Salvatore T. March
Institutions: University of Minnesota, University of Minnesota System
Language: English
Subfield: Information Systems
Field: Computer Science
Domain: Physical Sciences
Keywords: shared database, mathematical programming
Open Access status: bronze
OpenAlex ID: W2144811294
