Abstract
Abduction is a fundamental mode of reasoning with applications in many areas of AI and Computer Science. The computation of abductive explanations is an important computational problem, which is at the core of early systems such as the ATMS and Clause Management Systems and is intimately related to prime implicate generation in propositional logic. Many algorithms have been devised for computing some abductive explanation, and the complexity of the problem has been well studied. However, little attention has been paid to the problem of computing multiple explanations, and in particular all explanations for an abductive query. We fill this gap and consider the computation of all explanations of an abductive query from a propositional Horn theory, or of a polynomial subset of them. Our study pays particular attention to the form of the query, ranging from a literal to a compound formula, to whether explanations are based on a set of abducible literals and to the representation of the Horn theory, either by a Horn conjunctive normal form (CNF) or model-based in terms of its characteristic models. For these combinations, we present either tractability results in terms of polynomial total-time algorithms, intractability results in terms of nonexistence of such algorithms (unless P = NP), or semi-tractability results in terms of solvability in quasi-polynomial time, established by polynomial-time equivalence to the problem of dualizing a monotone CNF expression. Our results complement previous results in the literature, and refute a longstanding conjecture by Selman and Levesque. They elucidate the complexity of generating all abductive explanations and shed light on related problems such as generating sets of restricted prime implicates of a Horn theory. The algorithms for tractable cases can be readily applied for generating a polynomial subset of explanations in polynomial time.
- Angluin, D. 1987. Learning propositional Horn sentences with hints. Tech. Rep. Yale/DCS/RR-590, Dept. of Computer Science, Yale University.Google Scholar
- Apt, K., and Bezem, M. 1991. Acyclic programs. New Gen. Comput. 9, 3-4, 335--364.Google ScholarCross Ref
- Arimura, H. 1997. Learning acyclic first-order Horn sentences from entailment. In Proceedings 7th International Workshop on Algorithmic Learning Theory (ALT'97), M. Li and A. Maruoka, Eds. Lecture Notes in Computer Science, vol. 1316. Springer, 432--445. Google ScholarDigital Library
- Bioch, C., and Ibaraki, T. 1995. Complexity of identification and dualization of positive Boolean functions. Info. Comput. 123, 50--63. Google ScholarDigital Library
- Boros, E., Cepek, O., and Kogan, A. 1998. Horn minimization by iterative decomposition. Ann. Math. Artif. Intel. 23, 3-4, 321--343. Google ScholarDigital Library
- Boros, E., Crama, Y., and Hammer, P. L. 1990. Polynomial-time inference of all valid implications for Horn and related formulae. Ann. Math. Artif. Intel. 1, 21--32.Google ScholarDigital Library
- Brewka, G., Dix, J., and Konolige, K. 1997. Nonmonotonic Reasoning---An Overview. Number 73 in CSLI Lecture Notes. CSLI Publications, Stanford University, Stanford, CA.Google Scholar
- Bshouty, N. H. 1995. Exact learning Boolean functions via the monotone theory. Inf. Comput. 123, 146--153. Google ScholarDigital Library
- Bylander, T. 1991. The monotonic abduction problem: A functional characterization on the edge of tractability. In Proceedings of the 2nd International Conference on Principles of Knowledge Representation and Reasoning (KR-91). 70--77.Google Scholar
- Console, L., Sapino, M.L., and Theseider Dupré, D. 1995. The role of abduction in database view updating. J. Intel. Inf. Syst. 4, 3, 261--680. Google ScholarDigital Library
- Console, L., Theseider Dupré, D., and Torasso, P. 1991. On the relationship between abduction and deduction. J. Logic Comput. 1, 5, 661--690.Google ScholarCross Ref
- de Kleer, J. 1986. An assumption-based truth maintenance system. Artif. Intel. 28, 127--162. Google ScholarDigital Library
- Dechter, R., and Pearl, J. 1992. Structure identification in relational data. Artif. Intel. 58, 237--270. Google ScholarDigital Library
- del Val, A. 1999. A new method for consequence finding and compilation in restricted languages. In Proceedings of the 16th American National Conference on Artificial Intelligence (AAAI'99). AAAI Press/The MIT Press, Cambridge, MA, 259--264. Google ScholarDigital Library
- del Val, A. 2000a. On some tractable classes in deduction and abduction. Artif. Intel. 116, 1-2, 297--313. Google ScholarDigital Library
- del Val, A. 2000a. The complexity of restricted consequence finding and abduction. In Proceedings of the Seventeenth National Conference on Artificial Intelligence (AAAI-2000). AAAI Press/The MIT Press, Cambridge, MA, 337--342. Google ScholarDigital Library
- Dowling, W., and Gallier, J. H. 1984. Linear-time algorithms for testing the satisfiability of propositional Horn theories. J. Logic Prog. 3, 267--284.Google ScholarCross Ref
- Downey, R. G., Fellows, M., and Fellows, M. 1998. Parameterized Complexity. Springer-Verlag, New York.Google Scholar
- Eiter, T., and Gottlob, G. 1991. Identifying the minimal transversals of a hypergraph and related problems. Tech. Rep. CD-TR 91/16, Christian Doppler Laboratory for Expert Systems, TU Vienna, Austria. January. (Shortened version in SIAM J. Comput., 24(6):1278--1304, 1995). Google ScholarDigital Library
- Eiter, T., and Gottlob, G. 1995. The complexity of logic-based abduction. J. ACM 42, 1 (Jan.), 3--42. Google ScholarDigital Library
- Eiter, T., Gottlob, G., and Leone, N. 1997. Abduction from logic programs: Semantics and complexity. Theoret. Comput. Sci. 189, 1-2 (Dec.), 129--177. Google ScholarDigital Library
- Eiter, T., Gottlob, G., and Makino, K. 2002. New results on monotone dualization and generating hypergraph transversals. In Proceedings of the 34th ACM Symposium on Theory of Computing (STOC 2002). ACM, New York, 14--22. (Full paper SIAM J. Computing, 32(2):514--537, 2003.) Google ScholarDigital Library
- Eiter, T., and Makino, K. 2002. On computing all abductive explanations. In Proceedings of the 18th National Conference on Artificial Intelligence (AAAI '02). AAAI Press, 62--67. Google ScholarDigital Library
- Eiter, T., and Makino, K. 2003a. Abduction and the dualization problem. In Proceedings of the 6th International Conference on Discovery Science (DS 2003), Sapporo, Japan, Oct. 17--19, G. Grieser, Y. Tanaka, and A. Yamamoto, Eds. Lecture Notes in Computer Science/Lecture Notes in Artificial Intelligence, vol. 2843. Springer-Verlag, New York, 1--20.Google Scholar
- Eiter, T., and Makino, K. 2003b. Generating all abductive explanations for queries on propositional Horn theories. In Proceedings 12th Annual Conference of the EACSL (CSL 2003), (Vienna, Austria, Aug. 25--30) M. Baaz and J. Makowsky, Eds. Lecture Notes in Computer Science, vol. 2803, Springer-Verlag, New York, 197--211.Google Scholar
- Eiter, T., Makino, K., and Gottlob, G. 2007. Computational aspects of monotone dualization: A brief survey. Discrete Applied Mathematics, doi:10.1016/j.dam.2007.04.017. Google ScholarDigital Library
- Ekin, O., Hammer, P. L., and Kogan, A. 1999. On connected Boolean functions. Disc. Appl. Math. 96-97, 337--362. Google ScholarDigital Library
- Eshghi, K. 1993. A tractable class of abduction problems. In Proceedings IJCAI-93 (Chambery, France), Morgan Kaufmann, San Francisco, CA, 3--8.Google Scholar
- Fagin, R. 1982. Horn clauses and database dependencies. J. ACM 29, 952--985. Google ScholarDigital Library
- Fredman, M., and Khachiyan, L. 1996. On the complexity of dualization of monotone disjunctive normal forms. J. Algorithms 21, 618--628. Google ScholarDigital Library
- Friedrich, G., Gottlob, G., and Nejdl, W. 1990. Hypothesis classification, abductive diagnosis, and therapy. In Proceedings of the International Workshop on Expert Systems in Engineering. Lecture Notes in Artificial Intelligence, vol. 462, Springer-Verlag, New York, 69--78. Google ScholarDigital Library
- Ginsberg, M. L. 1988. Knowledge-base reduction: A new approach to checking knowledge bases for inconsistency & redundancy. In Proceedings of the 7th National Conference on Artificial Intelligence (AAAI-88). AAAI Press, 585--589.Google Scholar
- Goldsmith, J., and Sloan, R. 2005. New Horn revision algorithms. J. Mach. Learn. Res. 6, 1919--1938. Google ScholarDigital Library
- Goldsmith, J., Sloan, R. H., Szörényi, B., and Turán, G. 2004. New revision algorithms. In Proceedings 15th International Conference on Algorithmic Learning Theory (ALT 2004), S. Ben-David, J. Case, and A. Maruoka, Eds. Lecture Notes in Computer Science, vol. 3244. Springer-Verlag, New York, 395--409.Google ScholarCross Ref
- Gottlob, G., Pichler, R., and Wei, F. 2006. Bounded treewidth as a key to tractability of knowledge representation and reasoning. In Proceedings of the 21th National Conference on Artificial Intelligence (AAAI '06). AAAI Press. Google ScholarDigital Library
- Gottlob, G., Scarcello, F., and Sideri, M. 2002. Fixed parameter complexity in AI and nonmonotonic reasoning. Artif. Intel. 138, 1-2, 55--86. Google ScholarDigital Library
- Hammer, P., and Kogan, A. 1992. Horn functions and their DNFs. Inf. Proc. Lett. 44, 23--29. Google ScholarDigital Library
- Hammer, P., and Kogan, A. 1995. Quasi-acyclic propositional Horn knowledge bases: Optimal compression. IEEE Trans. Knowl. Data Engi. 7, 5, 751--762. Google ScholarDigital Library
- Hammer, P. L., and Kogan, A. 1994. Graph-based methods for Horn knowledge compression. In Proceedings of the 27th Hawaii International Conference on System Sciences, (HICSS'94). IEEE, 300--309.Google Scholar
- Ibaraki, T., Kogan, A., and Makino, K. 2003. Inferring minimal functional dependencies in Horn and q-Horn theories. Ann. Math. Artif. Intel. 38, 4, 233--255. Google ScholarDigital Library
- Inoue, K. 1992. Linear resolution for consequence finding. Artif. Intel. 56, 2-3, 301--354. Google ScholarDigital Library
- Johnson, D. S. 1990. A catalog of complexity classes. In Handbook of Theoretical Computer Science, J. van Leeuwen, Ed. Vol. A. Elsevier Science Publishers B.V. (North-Holland), Chap. 2. Google ScholarDigital Library
- Kakas, A., Kowalski, R., and Toni, F. 1998. The role of abduction in logic programming. In Handbook of Logic in Artificial Intelligence and Logic Programming, D. M. Gabbay, C. J. Hogger, and J. A. Robinson, Eds. Vol. V: Logic Programming. Clarendon Press, Oxford, UK, 235--324.Google Scholar
- Kautz, H., Kearns, M., and Selman, B. 1993. Reasoning with characteristic models. In Proceedings of the 11th National Conference on Artificial Intelligence (AAAI-93). AAAI Press, 34--39.Google Scholar
- Kautz, H., Kearns, M., and Selman, B. 1995. Horn approximations of empirical data. Artif. Intel. 74, 129--245. Google ScholarDigital Library
- Kavvadias, D., Papadimitriou, C., and Sideri, M. 1993. On Horn envelopes and hypergraph transversals. In Proceedings 4th International Symposium on Algorithms and Computation (ISAAC-93), W. Ng, Ed. Lecture Notes in Computer Science, vol. 762. Springer-Verlag, New York, 399--405. Google ScholarDigital Library
- Kavvadias, D. J., and Stavropoulos, E. C. 2003. Monotone Boolean dualization is in co-NP{log2n}. Inf. Proc. Lett. 85, 1--6. Google ScholarDigital Library
- Kean, A., and Tsiknis, G. 1990. An incremental method for generating prime implicants. J. Symb. Comput. 9, 185--206. Google ScholarDigital Library
- Kean, A., and Tsiknis, G. 1992a. Assumption based reasoning and clause management systems. Comput. Intell. 8, 1, 1--24.Google ScholarCross Ref
- Kean, A., and Tsiknis, G. 1992b. Clause management systems (CMS). Comput. Intell. 9, 1, 11--40.Google ScholarCross Ref
- Khardon, R. 1995. Translating between Horn representations and their characteristic Models. J. Artif. Intell. Res. 3, 349--372. Google ScholarDigital Library
- Khardon, R., Mannila, H., and Roth, D. 1999. Reasoning with examples: Propositional formulae and database dependencies. Acta Inform. 36, 4, 267--286.Google ScholarCross Ref
- Khardon, R., and Roth, D. 1996. Reasoning with models. Artif. Intell. 87, 1/2, 187--213. Google ScholarDigital Library
- Khardon, R., and Roth, D. 1997. Defaults and relevance in model-based reasoning. Artif. Intell. 97, 1/2, 169--193. Google ScholarDigital Library
- Levesque, H. 1986. Making believers out of computers. Artif. Intell. 30, 81--108. Google ScholarDigital Library
- Lonc, Z., and Truszczyński, M. 2003. Fixed-parameter complexity of semantics for logic programs. ACM Trans. Comput. Logic 4, 1, 91--119. Google ScholarDigital Library
- Lucchesi, C. L., and Osborn, S. 1978. Candidate keys for relations. J. Comput. Syst. Sci. 17, 270--279.Google ScholarCross Ref
- Makowsky, J. 1987. Why Horn formulas matter for computer science: initial structures and generic examples. J. Comput. Syst. Sci. 34, 2/3, 266--292. Google ScholarDigital Library
- Mannila, H., and Räihä, K.-J. 1986. Design by example: An application of Armstrong relations. J. Comput. Syst. Sci. 22, 2, 126--141. Google ScholarDigital Library
- Mannila, H., and Räihä, K.-J. 1988. Algorithms for inferring functional dependencies. Tech. Rep. A-1988-3, University of Tampere, Dept. of Comput Science, Series of Publ. A. April.Google Scholar
- Marquis, P. 2000. Consequence finding algorithms. In Handbook of Defeasible Reasoning and Uncertainty Management Systems, Vol. V: Algorithms for Uncertainty and Defeasible Reasoning. D. Gabbay and Ph.Smets, Eds. Kluwer Academic, 41--145.Google Scholar
- McKinsey, J. 1943. The decision problem for some classes of sentences without quantifiers. J. Symb. Logic 8, 61--76.Google ScholarCross Ref
- Nguyen, T. A., Perkins, W. A., Laffey, T. J., and Pecora, D. 1987. Knowledge-base verification. AI Magazine 8, 2, 69--75. Google ScholarDigital Library
- Papadimitriou, C. H. 1994. Computational Complexity. Addison-Wesley. Reading, MA.Google Scholar
- Poole, D. 1989. Explanation and prediction: An architecture for default and abductive reasoning. Comput. Intel. 5, 1, 97--110. Google ScholarDigital Library
- Provan, G. 1990. The computational complexity of multiple-context truth maintenance systems. In Proceedings of the ECAI-90. 522--527.Google Scholar
- Reddy, C., and Tadepalli, P. 1998. Learning first-order acyclic Horn programs from entailment. In Proceedings of the 15th International Conference on Machine Learning (ICML 1998), J. W. Shavlik, Ed. Morgan Kaufmann, San Francisco, CA, 472--480. Google ScholarDigital Library
- Reiter, R., and de Kleer, J. 1982. Foundations of assumption-based truth maintenance systems: Preliminary report. In Proceedings of the 6th National Conference on Artificial Intelligence (AAAI-87). AAAI Press, 183--188.Google Scholar
- Rutenburg, V. 1991. Complexity classification in truth maintenance systems. In Proceedings STACS-91. vol. 480, Lecture Notes in Computer Science. Springer-Verlag, New York, pp. 373--383. (Full paper In Annal. Math. Artif. Intel., 10:207--231, 1994.Google Scholar
- Selman, B., and Levesque, H. J. 1990. Abductive and default reasoning: A computational core. In Proceedings of the 8th National Conference on Artificial Intelligence (AAAI-90). AAAI Press, 343--348.Google Scholar
- Selman, B., and Levesque, H. J. 1996. Support set selection for abductive and default reasoning. Artif. Intell. 82, 259--272. Google ScholarDigital Library
- Valiant, L. G., and Vazirani, V. V. 1986. NP is as easy as detecting unique solutions. Theoret. Comput. Sci. 47, 1, 85--93. Google ScholarDigital Library
- Zanuttini, B. 2003. New polynomial classes for logic-based abduction. J. Artif. Intell. Res. 19, 1--10. Google ScholarDigital Library
Index Terms
- On computing all abductive explanations from a propositional Horn theory
Recommendations
Deductive inference for the interiors and exteriors of horn theories
In this article, we investigate deductive inference for interiors and exteriors of Horn knowledge bases, where interiors and exteriors were introduced by Makino and Ibaraki [1996] to study stability properties of knowledge bases. We present a linear ...
What makes propositional abduction tractable
Abduction is a fundamental form of nonmonotonic reasoning that aims at finding explanations for observed manifestations. This process underlies many applications, from car configuration to medical diagnosis. We study here the computational complexity of ...
A Complete Classification of the Complexity of Propositional Abduction
Abduction is the process of explaining a given query with respect to some background knowledge. For instance, $p$ is an explanation for the query $q$ given the knowledge $p\rightarrow q$. This problem is well known to have many applications, ...
Comments