Planning strategy representation in dolittle

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

This paper introduces multi-strategy planning and describes its implementation in the DOLITTLE system, which can combine many different planning strategies, including means-ends analysis, macro-based planning, abstraction-based planning (reduced and relaxed), and casebased planning on a single problem. Planning strategies are defined as methods to reduce the search space by exploiting some assumptions (socalled planning biases) about the problem domain. General operators are generalizations of standard STRIPS operators that conveniently represent many different planning strategies. The focus of this work is to develop a representation weak enough to represent a wide variety of different strategies, but still strong enough to emulate them. The search control method applies different general operators based on a strongest first principle; planning biases that are expected to lead to small search spaces are tried first. An empirical evaluation in three domains showed that multi-strategy planning performed significantly better than the best single strategy planners in these domains.

Original languageEnglish
Title of host publicationAdvances in Artificial Intelligence - 12th Biennial Conference of the Canadian Society for Computational Studies of Intelligence, AI 1998, Proceedings
EditorsRobert E. Mercer, Eric Neufeld
PublisherSpringer Verlag
Pages30-44
Number of pages15
ISBN (Print)3540645756, 9783540645757
DOIs
Publication statusPublished - 1998 Jan 1
Event12th Biennial Conference on Artificial Intelligence, AI 1998 - Vancouver, Canada
Duration: 1998 Jun 181998 Jun 20

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume1418
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other12th Biennial Conference on Artificial Intelligence, AI 1998
CountryCanada
CityVancouver
Period98/6/1898/6/20

Fingerprint

Planning
Search Space
Operator
Strategy
First-principles
Macros
Evaluation

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Baltes, J. (1998). Planning strategy representation in dolittle. In R. E. Mercer, & E. Neufeld (Eds.), Advances in Artificial Intelligence - 12th Biennial Conference of the Canadian Society for Computational Studies of Intelligence, AI 1998, Proceedings (pp. 30-44). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 1418). Springer Verlag. https://doi.org/10.1007/3-540-64575-6_38

Planning strategy representation in dolittle. / Baltes, Jacky.

Advances in Artificial Intelligence - 12th Biennial Conference of the Canadian Society for Computational Studies of Intelligence, AI 1998, Proceedings. ed. / Robert E. Mercer; Eric Neufeld . Springer Verlag, 1998. p. 30-44 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 1418).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Baltes, J 1998, Planning strategy representation in dolittle. in RE Mercer & E Neufeld (eds), Advances in Artificial Intelligence - 12th Biennial Conference of the Canadian Society for Computational Studies of Intelligence, AI 1998, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 1418, Springer Verlag, pp. 30-44, 12th Biennial Conference on Artificial Intelligence, AI 1998, Vancouver, Canada, 98/6/18. https://doi.org/10.1007/3-540-64575-6_38
Baltes J. Planning strategy representation in dolittle. In Mercer RE, Neufeld E, editors, Advances in Artificial Intelligence - 12th Biennial Conference of the Canadian Society for Computational Studies of Intelligence, AI 1998, Proceedings. Springer Verlag. 1998. p. 30-44. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/3-540-64575-6_38
Baltes, Jacky. / Planning strategy representation in dolittle. Advances in Artificial Intelligence - 12th Biennial Conference of the Canadian Society for Computational Studies of Intelligence, AI 1998, Proceedings. editor / Robert E. Mercer ; Eric Neufeld . Springer Verlag, 1998. pp. 30-44 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{5f04fc2a3194440da52af882b3a46490,
title = "Planning strategy representation in dolittle",
abstract = "This paper introduces multi-strategy planning and describes its implementation in the DOLITTLE system, which can combine many different planning strategies, including means-ends analysis, macro-based planning, abstraction-based planning (reduced and relaxed), and casebased planning on a single problem. Planning strategies are defined as methods to reduce the search space by exploiting some assumptions (socalled planning biases) about the problem domain. General operators are generalizations of standard STRIPS operators that conveniently represent many different planning strategies. The focus of this work is to develop a representation weak enough to represent a wide variety of different strategies, but still strong enough to emulate them. The search control method applies different general operators based on a strongest first principle; planning biases that are expected to lead to small search spaces are tried first. An empirical evaluation in three domains showed that multi-strategy planning performed significantly better than the best single strategy planners in these domains.",
author = "Jacky Baltes",
year = "1998",
month = "1",
day = "1",
doi = "10.1007/3-540-64575-6_38",
language = "English",
isbn = "3540645756",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "30--44",
editor = "Mercer, {Robert E.} and {Neufeld }, Eric",
booktitle = "Advances in Artificial Intelligence - 12th Biennial Conference of the Canadian Society for Computational Studies of Intelligence, AI 1998, Proceedings",

}

TY - GEN

T1 - Planning strategy representation in dolittle

AU - Baltes, Jacky

PY - 1998/1/1

Y1 - 1998/1/1

N2 - This paper introduces multi-strategy planning and describes its implementation in the DOLITTLE system, which can combine many different planning strategies, including means-ends analysis, macro-based planning, abstraction-based planning (reduced and relaxed), and casebased planning on a single problem. Planning strategies are defined as methods to reduce the search space by exploiting some assumptions (socalled planning biases) about the problem domain. General operators are generalizations of standard STRIPS operators that conveniently represent many different planning strategies. The focus of this work is to develop a representation weak enough to represent a wide variety of different strategies, but still strong enough to emulate them. The search control method applies different general operators based on a strongest first principle; planning biases that are expected to lead to small search spaces are tried first. An empirical evaluation in three domains showed that multi-strategy planning performed significantly better than the best single strategy planners in these domains.

AB - This paper introduces multi-strategy planning and describes its implementation in the DOLITTLE system, which can combine many different planning strategies, including means-ends analysis, macro-based planning, abstraction-based planning (reduced and relaxed), and casebased planning on a single problem. Planning strategies are defined as methods to reduce the search space by exploiting some assumptions (socalled planning biases) about the problem domain. General operators are generalizations of standard STRIPS operators that conveniently represent many different planning strategies. The focus of this work is to develop a representation weak enough to represent a wide variety of different strategies, but still strong enough to emulate them. The search control method applies different general operators based on a strongest first principle; planning biases that are expected to lead to small search spaces are tried first. An empirical evaluation in three domains showed that multi-strategy planning performed significantly better than the best single strategy planners in these domains.

UR - http://www.scopus.com/inward/record.url?scp=84958625097&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84958625097&partnerID=8YFLogxK

U2 - 10.1007/3-540-64575-6_38

DO - 10.1007/3-540-64575-6_38

M3 - Conference contribution

AN - SCOPUS:84958625097

SN - 3540645756

SN - 9783540645757

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 30

EP - 44

BT - Advances in Artificial Intelligence - 12th Biennial Conference of the Canadian Society for Computational Studies of Intelligence, AI 1998, Proceedings

A2 - Mercer, Robert E.

A2 - Neufeld , Eric

PB - Springer Verlag

ER -