An evaluation of explanations of probabilistic inference.

By H. J. Suermondt, Gregory F. Cooper. 1992

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Explanation in Human-AI Systems: A Literature Meta-Review, Synopsis of Key Ideas and Publications, and Bibliography for Explainable AI (2019, background): “(See also Antifakos et al., 2005; Dourish, 1995; Suermondt and Cooper, 1992; Tullio et al., 2007).”
A systematic review and taxonomy of explanations in decision support and recommender systems (2017, methodology): “Effectiveness Confidence+Sensor Data with Low Robot Ability [237] Decisive Input Values [213] Derived Topic Models and Time Intervals [199] Explanation Interfaces [183] Peer Graph Navigation [171] Tag-based [231] Transparency Confidence+Sensor Data with Low Robot Ability [237] Decisive Input Values ”
Graphical Explanation in Belief Networks (1997, background): “Chamberlain and Nordahl (1988), Cooper (1989), Druzdzel (1996), Henrion and Druzdzel (1990), Lauritzen and Spiegelhalter (1988), Pearl (1987), Suermondt (1991), and Suermondt and Cooper (1993) and have all made suggestions as to how such explanations might be generated.”
How it works: a field study of non-technical users interacting with an intelligent system (2007, methodology): “Suermondt and Cooper found that physicians using a medical expert system made fewer mistakes when provided with explanations of the system’s diagnoses [23].”
Explainable Artificial Intelligence: a Systematic Review (2020, methodology): “The first methods for explainability that were tested by human users are those generating textual explanations for Expert Systems (ES) in the 90’s [334, 335].”
Defining Explanation in Probabilistic Systems (1997, background): “Indeed, in experiments with medical diagnosis systems, medical students not only trusted the system more when presented with an explanation of the diagnosis, but also were more confident about disagreeing with it when the explanations did not account adequately for all of the aspects of the case (Su”
A review of explanation methods for Bayesian networks (2002, methodology): “Another method was Suermondt’s INSITE (Suermondt, 1992; Suermondt & Cooper, 1993), whose main objective is to identify the ﬁndings that inﬂuence the posterior probability of a certain hypothesis, as well as the paths through which the evidence ﬂows, which we will describe in Section 4.4.1.”