Efficient Satisfiability Verification for Conditional Importance Networks
- 530 Downloads
Conditional importance networks (CI-nets) provide a formal framework for modeling and reasoning with qualitative preferences over sets of many variables. Existing approaches for verifying the satisfiability of a CI-net operate on a complete model of the CI-net’s semantics, but the time required to construct this model limits the practical usefulness of CI-nets. We present an algorithm that decides a CI-net’s satisfiability by analyzing a sufficient partial model of its semantics, and we show how to efficiently construct such a model. Our method significantly reduces the average time needed to verify the satisfiability of CI-nets.
- 1.Bouveret, S., Endriss, U., Lang, J.: Conditional importance networks: a graphical language for representing ordinal, monotonic preferences over sets of goods. In: Boutilier, C. (ed.) Proceedings of the 21st International Joint Conference on Artificial Intelligence, IJCAI 2009, Pasadena, California, USA, 11–17 July 2009, pp. 67–72 (2009). http://ijcai.org/papers09/Papers/IJCAI09-022.pdf
- 2.Cornelio, C., Loreggia, A., Saraswat, V.A.: Logical conditional preference theories. CoRR abs/1504.06374 (2015). http://arxiv.org/abs/1504.06374
- 4.Santhanam, G.R., Basu, S., Honavar, V.: Dominance testing via model checking. In: Fox, M., Poole, D. (eds.) Proceedings of the 24th AAAI Conference on Artificial Intelligence, AAAI 2010, Atlanta, Georgia, USA, 11–15 July 2010. AAAI Press (2010). http://www.aaai.org/ocs/index.php/AAAI/AAAI10/paper/view/1844