Declarative local-search neighbourhoods in MiniZinc

Gustav Bjordal, Pierre Flener, Justin Pearson, Peter J. Stuckey, Guido Tack

Research output: Chapter in Book/Report/Conference proceedingConference PaperResearchpeer-review

Abstract

The aim of solver-independent modelling is to create a model of a satisfaction or optimisation problem independent of a particular technology. This avoids early commitment to a solving technology and allows easy comparison of technologies. MiniZinc is a solver-independent modelling language, supported by CP, MIP, SAT, SMT, and constraint-based local search (CBLS) backends. Some technologies, in particular CP and CBLS, require not only a model but also a search strategy. While backends for these technologies offer default search strategies, it is often beneficial to include in a model a user-specified search strategy for a particular technology, especially if the strategy can encapsulate knowledge about the problem structure. This is complex since a local-search strategy (comprising a neighbourhood, a heuristic, and a meta-heuristic) is often tightly tied to the model. Hence we wish to use the same language for specifying the model and the local search. We show how to extend MiniZinc so that one can attach a fully declarative neighbourhood specification to a model, while maintaining the solver-independence of the language. We explain how to integrate a model-specific declarative neighbourhood with an existing CBLS backend for MiniZinc.

Original languageEnglish
Title of host publicationProceedings - 2018 IEEE 30th International Conference on Tools with Artificial Intelligence, ICTAI 2018
Subtitle of host publication5–7 November 2018 Volos, Greece
EditorsMiltos Alamaniotis
Place of PublicationPiscataway NJ USA
PublisherIEEE, Institute of Electrical and Electronics Engineers
Pages98-105
Number of pages8
ISBN (Electronic)9781538674499
ISBN (Print)9781538674505
DOIs
Publication statusPublished - 2018
EventInternational Conference on Tools with Artificial Intelligence 2018 - Volos, Greece
Duration: 5 Nov 20187 Nov 2018
Conference number: 30th
https://web.archive.org/web/20180810120914/http://ictai2018.org/

Conference

ConferenceInternational Conference on Tools with Artificial Intelligence 2018
Abbreviated titleICTAI 2018
CountryGreece
CityVolos
Period5/11/187/11/18
Internet address

Keywords

  • (Constraint-based) local search
  • Declarative neighbourhood
  • Modelling

Cite this

Bjordal, G., Flener, P., Pearson, J., Stuckey, P. J., & Tack, G. (2018). Declarative local-search neighbourhoods in MiniZinc. In M. Alamaniotis (Ed.), Proceedings - 2018 IEEE 30th International Conference on Tools with Artificial Intelligence, ICTAI 2018: 5–7 November 2018 Volos, Greece (pp. 98-105). [8576023] Piscataway NJ USA: IEEE, Institute of Electrical and Electronics Engineers. https://doi.org/10.1109/ICTAI.2018.00025
Bjordal, Gustav ; Flener, Pierre ; Pearson, Justin ; Stuckey, Peter J. ; Tack, Guido. / Declarative local-search neighbourhoods in MiniZinc. Proceedings - 2018 IEEE 30th International Conference on Tools with Artificial Intelligence, ICTAI 2018: 5–7 November 2018 Volos, Greece. editor / Miltos Alamaniotis. Piscataway NJ USA : IEEE, Institute of Electrical and Electronics Engineers, 2018. pp. 98-105
@inproceedings{967f7203c7b940e5bbfc3172c34fc8fa,
title = "Declarative local-search neighbourhoods in MiniZinc",
abstract = "The aim of solver-independent modelling is to create a model of a satisfaction or optimisation problem independent of a particular technology. This avoids early commitment to a solving technology and allows easy comparison of technologies. MiniZinc is a solver-independent modelling language, supported by CP, MIP, SAT, SMT, and constraint-based local search (CBLS) backends. Some technologies, in particular CP and CBLS, require not only a model but also a search strategy. While backends for these technologies offer default search strategies, it is often beneficial to include in a model a user-specified search strategy for a particular technology, especially if the strategy can encapsulate knowledge about the problem structure. This is complex since a local-search strategy (comprising a neighbourhood, a heuristic, and a meta-heuristic) is often tightly tied to the model. Hence we wish to use the same language for specifying the model and the local search. We show how to extend MiniZinc so that one can attach a fully declarative neighbourhood specification to a model, while maintaining the solver-independence of the language. We explain how to integrate a model-specific declarative neighbourhood with an existing CBLS backend for MiniZinc.",
keywords = "(Constraint-based) local search, Declarative neighbourhood, Modelling",
author = "Gustav Bjordal and Pierre Flener and Justin Pearson and Stuckey, {Peter J.} and Guido Tack",
year = "2018",
doi = "10.1109/ICTAI.2018.00025",
language = "English",
isbn = "9781538674505",
pages = "98--105",
editor = "Miltos Alamaniotis",
booktitle = "Proceedings - 2018 IEEE 30th International Conference on Tools with Artificial Intelligence, ICTAI 2018",
publisher = "IEEE, Institute of Electrical and Electronics Engineers",
address = "United States of America",

}

Bjordal, G, Flener, P, Pearson, J, Stuckey, PJ & Tack, G 2018, Declarative local-search neighbourhoods in MiniZinc. in M Alamaniotis (ed.), Proceedings - 2018 IEEE 30th International Conference on Tools with Artificial Intelligence, ICTAI 2018: 5–7 November 2018 Volos, Greece., 8576023, IEEE, Institute of Electrical and Electronics Engineers, Piscataway NJ USA, pp. 98-105, International Conference on Tools with Artificial Intelligence 2018, Volos, Greece, 5/11/18. https://doi.org/10.1109/ICTAI.2018.00025

Declarative local-search neighbourhoods in MiniZinc. / Bjordal, Gustav; Flener, Pierre; Pearson, Justin; Stuckey, Peter J.; Tack, Guido.

Proceedings - 2018 IEEE 30th International Conference on Tools with Artificial Intelligence, ICTAI 2018: 5–7 November 2018 Volos, Greece. ed. / Miltos Alamaniotis. Piscataway NJ USA : IEEE, Institute of Electrical and Electronics Engineers, 2018. p. 98-105 8576023.

Research output: Chapter in Book/Report/Conference proceedingConference PaperResearchpeer-review

TY - GEN

T1 - Declarative local-search neighbourhoods in MiniZinc

AU - Bjordal, Gustav

AU - Flener, Pierre

AU - Pearson, Justin

AU - Stuckey, Peter J.

AU - Tack, Guido

PY - 2018

Y1 - 2018

N2 - The aim of solver-independent modelling is to create a model of a satisfaction or optimisation problem independent of a particular technology. This avoids early commitment to a solving technology and allows easy comparison of technologies. MiniZinc is a solver-independent modelling language, supported by CP, MIP, SAT, SMT, and constraint-based local search (CBLS) backends. Some technologies, in particular CP and CBLS, require not only a model but also a search strategy. While backends for these technologies offer default search strategies, it is often beneficial to include in a model a user-specified search strategy for a particular technology, especially if the strategy can encapsulate knowledge about the problem structure. This is complex since a local-search strategy (comprising a neighbourhood, a heuristic, and a meta-heuristic) is often tightly tied to the model. Hence we wish to use the same language for specifying the model and the local search. We show how to extend MiniZinc so that one can attach a fully declarative neighbourhood specification to a model, while maintaining the solver-independence of the language. We explain how to integrate a model-specific declarative neighbourhood with an existing CBLS backend for MiniZinc.

AB - The aim of solver-independent modelling is to create a model of a satisfaction or optimisation problem independent of a particular technology. This avoids early commitment to a solving technology and allows easy comparison of technologies. MiniZinc is a solver-independent modelling language, supported by CP, MIP, SAT, SMT, and constraint-based local search (CBLS) backends. Some technologies, in particular CP and CBLS, require not only a model but also a search strategy. While backends for these technologies offer default search strategies, it is often beneficial to include in a model a user-specified search strategy for a particular technology, especially if the strategy can encapsulate knowledge about the problem structure. This is complex since a local-search strategy (comprising a neighbourhood, a heuristic, and a meta-heuristic) is often tightly tied to the model. Hence we wish to use the same language for specifying the model and the local search. We show how to extend MiniZinc so that one can attach a fully declarative neighbourhood specification to a model, while maintaining the solver-independence of the language. We explain how to integrate a model-specific declarative neighbourhood with an existing CBLS backend for MiniZinc.

KW - (Constraint-based) local search

KW - Declarative neighbourhood

KW - Modelling

UR - http://www.scopus.com/inward/record.url?scp=85060826719&partnerID=8YFLogxK

U2 - 10.1109/ICTAI.2018.00025

DO - 10.1109/ICTAI.2018.00025

M3 - Conference Paper

SN - 9781538674505

SP - 98

EP - 105

BT - Proceedings - 2018 IEEE 30th International Conference on Tools with Artificial Intelligence, ICTAI 2018

A2 - Alamaniotis, Miltos

PB - IEEE, Institute of Electrical and Electronics Engineers

CY - Piscataway NJ USA

ER -

Bjordal G, Flener P, Pearson J, Stuckey PJ, Tack G. Declarative local-search neighbourhoods in MiniZinc. In Alamaniotis M, editor, Proceedings - 2018 IEEE 30th International Conference on Tools with Artificial Intelligence, ICTAI 2018: 5–7 November 2018 Volos, Greece. Piscataway NJ USA: IEEE, Institute of Electrical and Electronics Engineers. 2018. p. 98-105. 8576023 https://doi.org/10.1109/ICTAI.2018.00025