DATALOG¬ is a well-known logical query language, whose expressive power and data complexity under the stable model semantics has been recently determined. In this paper we consider the extension of DATALOG¬ to disjunctive DATALOG¬ (DDL¬), which allows disjunction in the head of program clauses, under the stable model semantics. We investigate and determine the expressiveness and the data complexity of DDL¬, as well as the expression complexity. The main findings of this paper are that disjunctive datalog captures precisely the class of all ∑2p-recognizable queries under the brave version of reasoning, and symmetrically the class of all Π2p-recognizable queries under the cautious version; the data complexity is ∑2p-completeness in the brave version, and Π2p-complete in the cautious version, while the expression complexity is NEXPTIMENP-complete in the brave version and co-NEXPTIMENP-complete in the cautious version.
CITATION STYLE
Eiter, T., Gottlob, G., & Mannila, H. (1994). Expressive power and complexity of disjunctive datalog under the stable model semantics. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 777 LNCS, pp. 83–103). Springer Verlag. https://doi.org/10.1007/3-540-57802-1_5
Mendeley helps you to discover research relevant for your work.