Abstract

Explanation is crucial in persuading others of the correctness of our beliefs to gain acceptance of our conclusion. Early research into expert systems focused on methods for reasoning. However it became apparent that the ultimate success lay in the ability to gain user acceptance by explaining the reasoning behind the conclusion. This study examines explanation from a social and an implementation standpoint. The social aspects of explanation provide insight into the role of naturally occurring explanations and listener expectations. Examination of research expert systems with explanation facilities and modifications of a simple Prolog expert system shell demonstrate the techniques required to simulate naturally occurring text. The modified shell produces improved explanations over the original shell, clearly indicating the desirability of natural appearance for gaining user acceptance.

Library of Congress Subject Headings

Expert systems (Computer science); Human-computer interaction; Explanation

Publication Date

1-1-1988

Document Type

Thesis

Student Type

Graduate

Degree Name

Computer Science (MS)

Department, Program, or Center

Computer Science (GCCIS)

Advisor

J. Al Biles

Advisor/Committee Member

Evelyn Rozanski

Advisor/Committee Member

Peter Anderson

Comments

Physical copy available from RIT's Wallace Library at QA76.76.E95 H443 1988

Campus

RIT – Main Campus

Download

Share

COinS