Lazy Arc Consistency

Arc consistency filtering is widely used in the framework of binary constraint satisfaction problems: with a low complexity, inconsistency may be detected,and domains are filtered. In this paper, we show that when detecting inconsistency is the objective, a systematic domain filtering is useless and a lazyapproach is more adequate. Whereas usual arc consistency algorithms produce the maximum arc consistent subdomain, when it exists, we propose a method, called LACY, which only looks for any arc consistent sub-domain. solution. When considering wipe-out detection only, the computation of the maximum arc consistent sub-domain is useless and one arc consistent sub-domain is sufficient since it proves the absence of wipe-out. In some cases, wipe-out detection alone is not enough: backtrack tree-search algorithms such as ReaZZyFUZZLookAhead or MAC also use domain sizes as a heuristic to choose the next variable to instantiate. Lazy arc consistency can then be extended to provide the additi...

• Type: Conference
• Year: 1996
• Where: AAAI
• Authors: Thomas Schiex, Jean-Charles Régin, Christine Gaspin, Gérard Verfaillie
• Keywords: AAAI 1996, Arc Consistency, Arc Consistent Sub-domain, Intelligent Agents, Maximum Arc