Evaluating Explanations

Psychology and philosophy have long studied the nature and role of explanation. More recently, artificial intelligence research has developed promising theories of how explanation facilitates learning and generalization. By using explanations to guide learning, explanation-based methods allow reliable learning of new concepts in complex situations, often from observing a single example. The author of this volume, however, argues that explanation-based learning research has neglected key issues in explanation construction and evaluation. By examining the issues in the context of a story understanding system that explains novel events in news stories, the author shows that the standard assumptions do not apply to complex real-world domains. An alternative theory is presented, one that demonstrates that context -- involving both explainer beliefs and goals -- is crucial in deciding an explanation's goodness and that a theory of the possible contexts can be used to determine which explanations are appropriate. This important view is demonstrated with examples of the performance of ACCEPTER, a computer system for story understanding, anomaly detection, and explanation evaluation.

First Published in 1994. Routledge is an imprint of Taylor & Francis, an informa company.

A new approach to the definition of scientific explanation. Unlike standard theories, it focuses initially on the explaining act itself, to which reference must be made in order to understand what an explanation is and how it can be evaluated in the sciences.

This report presents the conceptual foundations of the OECD Programme for International Student Assessment (PISA), now in its seventh cycle of comprehensive and rigorous international surveys of student knowledge, skills and well-being. Like previous cycles, the 2018 assessment covered reading, mathematics and science, with the major focus this cycle on reading literacy, plus an evaluation of students’ global competence – their ability to understand and appreciate the perspectives and world views of others. Financial literacy was also offered as an optional assessment.

Use of argumentation methods applied to legal reasoning is a relatively new field of study. The book provides a survey of the leading problems, and outlines how future research using argumentation-based methods show great promise of leading to useful solutions. The problems studied include not only these of argument evaluation and argument invention, but also analysis of specific kinds of evidence commonly used in law, like witness testimony, circumstantial evidence, forensic evidence and character evidence. New tools for analyzing these kinds of evidence are introduced.

This text provides a solid introduction to the foundations of research methods, with the goal of enabling students and professionals in the various fields of education to not simply become casual consumers of research who passively read bits and pieces of research articles, but discerning consumers able to effectively use published research for practical purposes in educational settings. All issues important for understanding and using published research for these purposes are covered. Key principles are illustrated with research studies published in refereed journals across a wide spectrum of education. Exercises distributed throughout the text encourage readers to engage interactively with what they are reading at the point when the information is fresh in their minds. This text is designed for higher level undergraduate and graduate programs. Course instructors will find that it provides a solid framework in which to promote student interaction and discussion on important issues in research methodology.

The first comprehensive exploration of the nature and value of understanding, addressing burgeoning debates in epistemology and philosophy of science.

Psychology and philosophy have long studied the nature and role of explanation. More recently, artificial intelligence research has developed promising theories of how explanation facilitates learning and generalization. By using explanations to guide learning, explanation-based methods allow reliable learning of new concepts in complex situations, often from observing a single example. The author of this volume, however, argues that explanation-based learning research has neglected key issues in explanation construction and evaluation. By examining the issues in the context of a story understanding system that explains novel events in news stories, the author shows that the standard assumptions do not apply to complex real-world domains. An alternative theory is presented, one that demonstrates that context -- involving both explainer beliefs and goals -- is crucial in deciding an explanation's goodness and that a theory of the possible contexts can be used to determine which explanations are appropriate. This important view is demonstrated with examples of the performance of ACCEPTER, a computer system for story understanding, anomaly detection, and explanation evaluation.