Computing scenario from knowledge with preferentially ordered hypotheses

Abstract

SUMMARYA method of computing scenario for knowledge including preferentially ordered hypotheses is proposed. Minimum conflict sets are found for competing hypothesis sets, and acceptable hypothesis sets are then found by applying preference relation to the minimum conflict sets; the disjunction of all elements of the acceptable hypothesis set thus obtained can be added to the fact set without contradiction. A fact set combined with all such disjunctions is considered a scenario, and is used for reasoning. Basically, the acceptability of hypothesis subsets for their minimal conflict sets is to be checked in order to obtain a scenario, which requires combinatorial calculations. A method is proposed to simplify such calculations; namely, acceptability is first checked for the single-element hypothesis sets, and then set operations are employed to check the acceptability of an arbitrary hypothesis set.

Autonomous Hypothetical Reasoning: the Case for Open-minded Agents

“…The proposed logic is firstorder logic enhanced with special notation for the representation of Reiter's original default rules and for the derivation of extensions. Tahara in (Tahara, 2004) addresses the issue of inconsistency that may arises in the knowledge base as a result of inconsistent hypotheses and uses a preference ordering in order to resolve contradictions.…”

Three‐valued logic for reasoning from an inconsistent knowledge base

“…(1) is based on classical logic, and has two major methods: (1a) deductive reasoning under consistent maximal subsets chosen from the inconsistent knowledge base [1][2][3], and (1b) deriving acceptable consequences based on the relationship between arguments which are pairs of consequences and consistent minimal subsets which derive the consequences [4][5][6][7]. Both methods have the same problem in the condition to assure the validity of the derived conclusions, since maximal consistent sets or arguments are contradictory to each other.…”

Reasoning with inconsistent knowledge base

“…(1) A method based on consistency: deductive reasoning under maximal consistent sets selected from an inconsistent knowledge base [1][2][3] (2) A method based on argumentation: selection of acceptable conclusions according to the conflict relation between arguments, which are pairs of a conclusion and its support (a minimal consistent knowledge base) [4,5] (3) A method based on multivalued logic (3-valued logic): reasoning according to a multivalued logic semantics [6,7] In this paper we concentrate on binary logic, and consider methods (1) and (2).…”