Journal of Automated Reasoning

, Volume 2, Issue 2, pp 109–126 | Cite as

Correctness criteria of some algorithms for uncertain reasoning using Incidence Calculus

  • Alan Bundy


Incidence Calculus is a technique for associating uncertainty values with logical sentences. These uncertainty values are called incidences and they are sets of points, which may be thought of as representing equivalence classes of situations, Tarskian models, or possible worlds. Incidence Calculus was originally introduced in [1].

Incidence Calculus was designed to overcome various inherent problems with purely numeric mechanisms for uncertain reasoning [2]. In particular, incidences can represent the dependence between sentences, which numbers cannot, and hence Incidence Calculus can provide genuine, probabilistic reasoning.

In this paper we prove soundness and completeness results for some algorithms introduced in [1] and hence satisfy some of the correctness criteria for Incidence Calculus. These algorithms can be used for probabilistic reasoning and to check the consistency of the subjective probabilities of sentences.

Key words

Incidence Calculus probability uncertainty inference logic expert systems 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    BundyA., ‘Incidence Calculus: A Mechanism for Probabilistic Reasoning”,J. Automated Reasoning,1, 263–284 (1985).zbMATHMathSciNetCrossRefGoogle Scholar
  2. 2.
    White, A. P., ‘Inference deficiencies in rule-based expert systems’, inResearch and Development in Expert Systems (ed. M. A. Bramer), Cambridge University Press, pp. 39–50 (1984).Google Scholar

Copyright information

© D. Reidel Publishing Company 1986

Authors and Affiliations

  • Alan Bundy
    • 1
  1. 1.Department of Artificial IntelligenceUniversity of EdinburghEdinburghScotland

Personalised recommendations