Parallel Query Optimization: Exploiting Bushy and Pipeline Parallelism with Genetic Programs
Created by W.Langdon from
gp-bibliography.bib Revision:1.8081
- @TechReport{stillger:1996:tr,
-
author = "Michael Stillger and Myra Spiliopoulou and
Johann-Christoph Freytag",
-
title = "Parallel Query Optimization: Exploiting Bushy and
Pipeline Parallelism with Genetic Programs",
-
institution = "DBIS, Humboldt University",
-
year = "1996",
-
type = "Technical Report",
-
number = "HUB-IB-65",
-
address = "Berlin, Germany",
-
month = jun,
-
keywords = "genetic algorithms, genetic programming",
-
URL = "http://www.dbis.informatik.hu-berlin.de/fileadmin/research/papers/techreports/1996-hub_ib_65-stillger.ps.gz",
-
URL = "http://citeseer.ist.psu.edu/stillger96parallel.html",
-
abstract = "Parallel query optimization is one of the hardest
problems in the databases area. The various cost models
reflecting the query execution parameters determine the
structure and size of the solutions space. To explore
this space, research has turned towards combinatorial
optimization techniques, heuristics and genetic
algorithms, which have been primarily studied for
sequential query processing. In this study, we propose
a genetic programming strategy for the optimization of
parallel bushy query execution plans. Genetic
programming has evolved from genetic algorithms, and is
more flexible and expressive. We consider two cost
functions modelling different modes of interoperator
parallelism. We analyse the behaviour of the search
strategy and observe that it is affected by the cost
function. Despite this, our experiments show that our
strategy converges to optimal plans of very good
quality, and performs best when bushy interoperator
parallelism is exploited.",
-
size = "33 pages",
- }
Genetic Programming entries for
Michael Stillger
Myra Spiliopoulou
Johann-Christoph Freytag
Citations