TY - JOUR
T1 - Generating diverse plans to handle unknown and partially known user preferences
AU - Nguyen, Tuan Anh
AU - Do, Minh
AU - Gerevini, Alfonso Emilio
AU - Serina, Ivan
AU - Srivastava, Biplav
AU - Kambhampati, Subbarao
N1 - Funding Information:
We thank Menkes van den Briel for drawing our attention to ICP measure initially. Kambhampati’s research is supported in part by an IBM Faculty Award, the NSF grant IIS2013308139, ONR grants N00014-09-1-0017, N00014-07-1-1049, N000140610058, and by a Lockheed Martin subcontract TT0687680 to ASU as part of the DARPA Integrated Learning program. Tuan Nguyen was also supported by a Science Foundation of Arizona fellowship.
PY - 2012/10
Y1 - 2012/10
N2 - Current work in planning with preferences assumes that user preferences are completely specified, and aims to search for a single solution plan to satisfy these. In many real world planning scenarios, however, the user may provide no knowledge or at best partial knowledge of her preferences with respect to a desired plan. In such situations, rather than presenting a single plan as the solution, the planner must instead provide a set of plans containing one or more plans that are similar to the one that the user really prefers. In this paper, we first propose the usage of different measures to capture the quality of such plan sets. These are domain-independent distance measures based on plan elements (such as actions, states, or causal links) if no knowledge of the user preferences is given, or the Integrated Convex Preference (ICP) measure in case incomplete knowledge of such preferences is provided. We then investigate various heuristic approaches to generate sets of plans in accordance with these measures, and present empirical results that demonstrate the promise of our methods.
AB - Current work in planning with preferences assumes that user preferences are completely specified, and aims to search for a single solution plan to satisfy these. In many real world planning scenarios, however, the user may provide no knowledge or at best partial knowledge of her preferences with respect to a desired plan. In such situations, rather than presenting a single plan as the solution, the planner must instead provide a set of plans containing one or more plans that are similar to the one that the user really prefers. In this paper, we first propose the usage of different measures to capture the quality of such plan sets. These are domain-independent distance measures based on plan elements (such as actions, states, or causal links) if no knowledge of the user preferences is given, or the Integrated Convex Preference (ICP) measure in case incomplete knowledge of such preferences is provided. We then investigate various heuristic approaches to generate sets of plans in accordance with these measures, and present empirical results that demonstrate the promise of our methods.
KW - Diverse plans
KW - Heuristics
KW - Partial preferences
KW - Planning
KW - Search
UR - http://www.scopus.com/inward/record.url?scp=84863479544&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84863479544&partnerID=8YFLogxK
U2 - 10.1016/j.artint.2012.05.005
DO - 10.1016/j.artint.2012.05.005
M3 - Article
AN - SCOPUS:84863479544
SN - 0004-3702
VL - 190
SP - 1
EP - 31
JO - Artificial Intelligence
JF - Artificial Intelligence
ER -