Method for partitioning multi-dimensional data sets into rectangular partitions
AssigneeLucent Technologies Inc.
InventorsShanmuugavelayut Muthukrishnan, Viswanath Poosala, Torsten Suel
Life of this patent
Estimating the term…
Abstract
A method of using an algorithm to partition a multi-dimensional data set into a minimum number of rectangular-shaped partitions that can be processed more quickly than the non-partitioned data set while satisfying certain specified performance constraints.
Claims20
A method for partitioning a multidimensional data set, comprising the steps of:
- selecting a metric for evaluating a partitioning of a data set;
- selecting a performance constraint which the metric may not exceed;
- constructing an n-dimensional partition from the data set;
- and
- running an algorithm to determine the partitioning of the data set, wherein the metric does not exceed the performance constraint thereby ensuring that the data set is partitioned into a minimum number of partitions which enable the data in the data set to be processed more quickly than the data set.
A method for partitioning a two dimensional data set into a plurality of partitions which can be processed more quickly than the two dimensional data set from which they are derived, comprising the steps of:
- selecting a performance constraint which a metric used to evaluate the partitioning may not exceed;
- constructing an n-dimensional partition from the data set;
- and
- running an algorithm to determine the partitioning of the data set, wherein the data set is partitioned into a minimum number of partitions, thereby enabling the partitioning to be processed more quickly than the data set from which it is derived.
A method for determining the minimum number of tiles into which an n-dimensional data set is to be partitioned for processing the data set as quickly as possible, comprising the steps of:
- selecting a performance constraint which a metric used to evaluate a partitioning of a data set may not exceed;
- and
- running an algorithm to determine the partitioning of the data set, wherein the data set is partitioned into a minimum number of partitions, thereby enabling the partitioned data to be processed more quickly than the data set.
Dependencies are inferred from the claim text.