Construction of optimal indexes

Abstract

An algorithm to design index structures for a data base is presented. This index, instead of partitioning a data base based on attribute-value pairs, effects a partition based on the nature of the user queries. A significant feature of the algorithm is that instead of working with the actual attribute values in a record, it operates on a hash vector… (More)
DOI: 10.1145/1795396.1795447

Topics

  • Presentations referencing similar topics