Year: 2014
Type: article
Abstract: Abduction, or inference to the best explanation, is, plausibly, part of commonsense reasoning, and a means by which a cognitive system may arrive at estimates of its world from observational and other evidence.We take this "world estimate" to be the cognitive system's beliefs.Since such reasoning is fallible, and world estimates will sometimes contain errors, an abductive reasoning system might improve its performance if it has a way to engage in belief revision when new evidence, or further reasoning, indicates the existence of a problem.In this study, we develop, implement, and experimentally validate a metareasoning system that monitors and attempts to correct beliefs established by the base-level abductive reasoning system.We first identify that the presence of an anomaly, which we define as an observation or other evidence that cannot plausibly and consistently be explained, as a signal that the cognitive system's world estimate might be incorrect or, alternatively, that the unexplainable datum is noise.The metareasoning system responds to the presence of anomalies by asking exactly that question: which anomalies are due to mistakes in the world estimate, and warrant specific belief revisions, and which anomalies are due to noise, and should not instigate belief revisions?Various considerations regarding the nature of the anomalies and the system's reasoning history are brought to bear to answer this question.Fundamentally, we see the metareasoning question ("what explains these anomalies: mistaken beliefs, or noise?") as structurally similar to the cognitive system's original question, "what explains these observations?"Thus, the metareasoning system is an abductive reasoning system, just like the base-level system.The anomalies constitute meta-evidence which may be explained by meta-hypotheses.These meta-hypotheses describe the various was, "Well, it looks like we're karmically stuck to work together."I have never regretted connecting with JJ nor have I been jealous of the attention and assistance given to other graduate students from their advisors.In fact, I feel incredibly lucky (karmically or otherwise) to have been able to work with someone who is so deeply sophisticated in the issues relating to this thesis.I am continually inspired by the direct and honest approach he takes to his work and student advising.I recall asking him once if he had any advice for how I should write a certain research article.He said, "Be as clear as possible."That phrase has stuck with me like a zen koan and now guides all of my writing, teaching, and thinking.I want to thank JJ for being as clear as possible.I also want to thank B. Chandrasekaran (Chandra) and Neil Tennant for serving on my committee and providing guidance over the years.Both Chandra and Neil bring an incredible depth of knowledge ensuring that every conversation gives me a new perspective on my work.They both are experts at finding the deeper meanings of surface-level phenomena and dissecting and later synthesizing the various efforts of other researchers.I also have benefited from their close inspection of the present work though, of course, all muddied arguments, ad hoc designs, and weak analyses remain my own.My undergraduate advisor, Guy-Alain Amoussou, also deserves acknowledgment for his forethoughtful insistence that I participate in undergraduate research and teaching.At the time, I did not attribute any particular long-term value on my participation in these v programs.However, I now believe that Guy-Alain had a grand plan for me and the other students he advised.I attribute my entrance and success in graduate school to the work we did together.And six years after I finished my undergraduate education, I found myself discussing those activities in job interviews.Guy-Alain's guidance so long ago helped me secure an assistant professor position, and I would not be surprised if that was his plan all along.I would also like to thank Norsys Software Corp. (www.norsys.com)for their development and support of Netica, for which they kindly provided me a gratis license.The Netica library was used extensively in this work for efficient probabilistic reasoning with Bayesian networks.Finally, I want to acknowledge the unbounded support from my wife, Tracy Dawson, and father and step-mom, John Eckroth and Kathryn Golden.My parents systematically removed every obstacle barring my success (except my own laziness, of course), while my wife was capable even of stamping out my laziness (well, sometimes).She has always seen the light at the end of the tunnel and, if I am able to return the favor, the destination beyond the tunnel will be very bright, sunny, and warm indeed.
Author: Joshua Eckroth
Institution: The Ohio State University
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W2521690325