A popular approach to explanations amounts to backward chaining over logical implications encoding causal links. However, the resulting explanations are often unsatisfactory from a common-sense point of view. We dene a framework allowing us to distinguish causal implication from mere logical implication. Causal explanations are then deduced through two inference schemes so that explaining is in some way "less than implying" and "more than implying". Finally, we show how our approach applies to diagnostics.
CITATION STYLE
Besnard, P., & Cordier, M. O. (1999). Inferring causal explanations. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 1638, pp. 55–67). Springer Verlag. https://doi.org/10.1007/3-540-48747-6_6
Mendeley helps you to discover research relevant for your work.