On the Semantics of Logic Programs with Preferences

This work is a contribution to realizing prioritized reasoning in logic programming in the presence of preference relations involving atoms. In more details, the case of dynamic preferences is investigated and a semantics interpreting each preference rule as a tool for representing a choice over alternative options is proposed. The technique, providing a new interpretation for prioritized logic programs, is inspired by the one proposed by Sakama and Inoue in [19] and enriched with the use of structural information of preference rules as proposed by Brewka et al. in [6]. Specifically, the analysis of the logic program is carried out together with the analysis of preferences in order to determine the choice order and the sets of comparable models. The proposed approach is compared with those in [6,19]. Complexity analysis is also performed showing that the use of additional information does not increase the complexity of computing preferred stable models.