Query-Driven Coordination of Multiple Answer Sets

Abstract

This article studies coordination protocols between logical agents to answer queries to different types of combined programs. More precisely, we consider a system of agents corresponding to different logic programs under the answer set semantics, and different kind of coordination semantics to combine them: generous coordination (gathering all the answer sets of all agents), rigorous coordination (selecting answer sets shared by all agents), composition (building consistent union of answer sets from each agent) and consensus (taking intersection of answer sets from each agent). Rather than explicitly building a coordination program, which would require to compute all answer sets of each agent, we propose to use coordination protocols that would only compute answer sets that are needed to answer a query to the coordination. In this paper, after presenting our context and the coordination semantics we are using, we define coordination protocols for answering queries on them, translating constraints on the coordination into local constraints. Some examples are then given to illustrate the expressiveness of these basic types of coordination when combined together and possible applications are discussed.