TY - GEN
T1 - Multi-agent action modeling through action sequences and perspective fluents
AU - Baral, Chitta
AU - Gelfond, Gregory
AU - Pontelli, Enrico
AU - Cao Son, Iran
N1 - Publisher Copyright:
© Copyright 2015, Association for the Advancement of Artificial Intelligence. All rights reserved.
PY - 2015
Y1 - 2015
N2 - Actions in a multi-agent setting have complex characteristics. They may not only affect the real world, but also affect the knowledge and beliefs of agents in the world. In many cases, the effect on the beliefs or knowledge of an agent is not due to that agent actively doing some actions, but could be simply the result of that agent's perspective in terms of where it is looking. In dynamic epistemic logic (DEL), such multi-agent actions are expressed as complex constructs or as Kripke model type structures. This paper uses the multi-agent action language mA+ to show how one can take advantage of some of the perspective fluents of the world to model complex actions, in the sense of DEL, as simple action sequences. The paper describes several plan modules using such actions. Such plan modules will be helpful in planning for belief and knowledge goals in a multi-agent setting, as planning from scratch would often be prohibitively time consuming.
AB - Actions in a multi-agent setting have complex characteristics. They may not only affect the real world, but also affect the knowledge and beliefs of agents in the world. In many cases, the effect on the beliefs or knowledge of an agent is not due to that agent actively doing some actions, but could be simply the result of that agent's perspective in terms of where it is looking. In dynamic epistemic logic (DEL), such multi-agent actions are expressed as complex constructs or as Kripke model type structures. This paper uses the multi-agent action language mA+ to show how one can take advantage of some of the perspective fluents of the world to model complex actions, in the sense of DEL, as simple action sequences. The paper describes several plan modules using such actions. Such plan modules will be helpful in planning for belief and knowledge goals in a multi-agent setting, as planning from scratch would often be prohibitively time consuming.
UR - http://www.scopus.com/inward/record.url?scp=84986877408&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84986877408&partnerID=8YFLogxK
M3 - Conference contribution
AN - SCOPUS:84986877408
T3 - AAAI Spring Symposium - Technical Report
SP - 25
EP - 31
BT - Logical Formalizations of Commonsense Reasoning - Papers from the AAAI Spring Symposium, Technical Report
PB - AI Access Foundation
T2 - 2015 AAAI Spring Symposium
Y2 - 23 March 2015 through 25 March 2015
ER -