Abstract
We introduce PCP-nets, a formalism to model qualitative conditional preferences with probabilistic uncertainty. PCP-nets generalise CP-nets by allowing for uncertainty over the preference orderings. We define and study both optimality and dominance queries in PCP-nets, and we propose a tractable approximation of dominance which we show to be very accurate in our experimental setting. Since PCP-nets can be seen as a way to model a collection of weighted CP-nets, we also explore the use of PCP-nets in a multi-agent context, where individual agents submit CP-nets which are then aggregated into a single PCP-net. We consider various ways to perform such aggregation and we compare them via two notions of scores, based on well known voting theory concepts. Experimental results allow us to identify the aggregation method that better represents the given set of CP-nets and the most efficient dominance procedure to be used in the multi-agent context.
Original language | English |
---|---|
Pages (from-to) | 1103-1161 |
Number of pages | 59 |
Journal | Journal of Artificial Intelligence Research |
Volume | 72 |
DOIs | |
State | Published - 2021 |
Bibliographical note
Funding Information:Judy Goldsmith’s work was partially supported by NSF grant IIS-1649152. The work of Cristina Cornelio and Francesca Rossi was partially supported by the University of Padova project “Incorporating patients’ preferences in kidney transplant decision protocols”. The work of Cristina Cor-nelio was partially done while at IBM Research. Nicholas Mattei’s work was partially supported by NICTA, funded by the Australian Government through the Department of Communications and the Australian Research Council through the ICT Centre of Excellence Program. The work of Nicholas Mattei was also partially done while at IBM Research. We would also like to thank the anonymous referees whose careful feedback made this a better paper than it had been.
Publisher Copyright:
© 2021 AI Access Foundation. All rights reserved.
ASJC Scopus subject areas
- Artificial Intelligence