Controlling search in constrained-object models

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

6 Scopus citations

Abstract

A constrained-object model is a collection of classes parameterized by constraints and connected through composition and inheritance relations. A class is classically a factory of objects, that correctly linked are able to smoothly capture the inherent structure of complex and large constraint-based problems. Constrained-object languages are mainly focused on the definition of object-oriented, elegant, concise and expressive models. However, little attention has been paid in defining primitives to specify the search strategy of the solving process. In this paper, we address this concern by introducing an extensible formalism to define search strategies involving multiple options. Our main goal is to allow the user to customize, improve, and analyze the search process of constrained-object models.

Original languageEnglish
Title of host publicationAdvances in Artificial Intelligence, IBERAMIA 2010 - 12th Ibero-American Conference on AI, Proceedings
Pages582-591
Number of pages10
DOIs
StatePublished - 16 Dec 2010
Event12th Ibero-American Conference on Advances in Artificial Intelligence, AI 2010 - Bahia Blanca, Argentina
Duration: 1 Nov 20105 Nov 2010

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume6433 LNAI
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference12th Ibero-American Conference on Advances in Artificial Intelligence, AI 2010
CountryArgentina
CityBahia Blanca
Period1/11/105/11/10

Keywords

  • Constraint Satisfaction
  • Search

Fingerprint Dive into the research topics of 'Controlling search in constrained-object models'. Together they form a unique fingerprint.

Cite this