Generating explanations in context: The system perspective - ScienceDirect

Expert Systems with Applications

Volume 8, Issue 4, 1995, Pages 491-503

Expert Systems with Applications

https://doi.org/10.1016/0957-4174(94)E0038-V Get rights and content

Abstract

Explanations for expert systems are best provided in context, and, recently, many systems have used some notion of context in different ways in their explanation module. For example, some explanation systems take into account a user model. Others generate an explanation depending on the preceding and current discourse. In this article, we bring together these different notions of context as elements of a global picture that might be taken into account by an explanation module, depending on the needs of the application and the user. We characterize each of these elements, describe the constraints they place on communication, and present examples to illustrate the points being made. We discuss the implications of these different aspects of context on the design of explanation facilities. Finally, we describe and illustrate with examples, an implemented intention-based planning framework for explanation that can take into account the different aspects of context discussed above.

References (68)

J.F. Allen et al.
Analyzing intention in utterances
Artificial Intelligence
(1980)
R.B. Allen
User models: Theory, method, and practice
International Journal of Man-Machine Studies
(1990)
S.J. Kaplan
Cooperative responses from a portable natural language query system
Artificial Intelligence
(1982)
K.F. McCoy
Generating context sensitive responses to object-related misconceptions
Artificial Intelligence
(1989)
D.E. Appelt
Planning English sentences
(1985)
J.A. Bateman et al.
Phrasing a text in terms the user can understand
D. Biber
A typology of English texts
Linguistics
(1989)
J. Bowers et al.
Local and global structuring of computer mediated communication: Developing linguistic perspectives on CSCW in cosmos
P.J. Brézillon
B. Cahour
Competence modelling in consultation dialogs

B. Cahour et al.

Role and use of user models

S. Carberry

Plan recognition in natural language dialogue

(1990)

G. Carenini et al.

Generating explanations in context

A. Cawsey

Planning interactive explanations

International Journal of Man-Machine Studies

(1993)

B. Chandrasekaran et al.

Explanations in knowledge systems: The role of explicit representation of design knowledge

IEEE Expert

(1991)

P.R. Cohen et al.

Rational interaction as the basis for communication

R. de Beaugrande

Text, discourse and process

(1980)

DEC

Network training solutions: Introduction to data communications—Student guide

(1987)

DEC

M. Elhadad et al.

Generating customized text and graphics in the COMET explanation testbed

S. Feiner et al.

Automating the generation of coordinated multimedia explanations

IEEE Computer

(1991)

B.A. Goodman et al.

On the interaction between plan recognition and intelligent interfaces

User Modeling and User-Adapted Interaction

(1992)

M. Gregory et al.

Language and Situation

(1978)

B.J. Grosz et al.

Attention, intention, and the structure of discourse

Computational Linguistics

(1986)

M.A.K. Halliday

language as social semiotic

(1978)

M.A.K. Halliday

Language as code and language as behavior: A systemic functional interpretation of the nature and ontogenesis of dialogue

R. Hasan

What's going on: A dynamic view of context

J.R. Hobbs et al.

Why ask?

E.H. Hovy et al.

Employing knowledge resources in a new text planner architecture

S. Kerpedjiev

Automatic generation of multimodal weather reports from datasets

R. Kittredge et al.

On the need for domain communication knowledge

Computational Intelligence

(1991)

A. Kobsa et al.

User models in dialog systems

A. Kobsa et al.

Computational linguistics

(1988)

K. Kukich

Knowledge based report generation: A knowledge-engineering approach to natural language report generation

Cited by (23)

Evaluating XAI: A comparison of rule-based and example-based explanations
2021, Artificial Intelligence
Citation Excerpt :
The second set of recommendations regards the experimental context, including the use case. The use case determines the task, the participants that can and should be used, the mode of the interaction, the communication that takes place and the information available to the user [53]. As [4] already stated, the selected use case has a large effect on the conclusions that can be drawn and the extent to which they can be generalized.
Current developments in Artificial Intelligence (AI) led to a resurgence of Explainable AI (XAI). New methods are being researched to obtain information from AI systems in order to generate explanations for their output. However, there is an overall lack of valid and reliable evaluations of the effects on users' experience of, and behavior in response to explanations. New XAI methods are often based on an intuitive notion what an effective explanation should be. Rule- and example-based contrastive explanations are two exemplary explanation styles. In this study we evaluate the effects of these two explanation styles on system understanding, persuasive power and task performance in the context of decision support in diabetes self-management. Furthermore, we provide three sets of recommendations based on our experience designing this evaluation to help improve future evaluations. Our results show that rule-based explanations have a small positive effect on system understanding, whereas both rule- and example-based explanations seem to persuade users in following the advice even when incorrect. Neither explanation improves task performance compared to no explanation. This can be explained by the fact that both explanation styles only provide details relevant for a single decision, not the underlying rational or causality. These results show the importance of user evaluations in assessing the current assumptions and intuitions on effective explanations.
Semantic network representation of computer-mediated discussions: Conceptual facilitation form and knowledge acquisition
2008, Omega
Prior research on computer-mediated discussions examined their effects on knowledge acquisition without considering the role of the discussion representation. In this study, we investigate and compare the effect of semantic network discussion representations on knowledge acquisition to that of the threaded representations featured in most traditional discussion forums. Furthermore, we identify, define and operationalize a new, i.e., conceptual facilitation (validation of the conceptual organization of the discussion), assessing its role in knowledge acquisition at different levels of restrictiveness. The empirical results of a field experiment indicate that semantic network discussion representations enable the acquisition of more complex and better-integrated knowledge structures than threaded discussion representations. Conceptual facilitation forms entailing different levels of restrictiveness are also found to play a significant role. As the level of conceptual facilitation restrictiveness decreases, knowledge acquisition improves. Our findings empirically demonstrate the importance of accounting for discussion representation as a contingency factor in explaining group discussion processes and outcomes. To practitioners, our study provides empirical evidence on the advantages of semantic networks over threaded representations as an alternative mode for computer-mediated discussion representations. We also suggest guidelines for the selection of appropriate conceptual facilitation for discussion forums intended for knowledge acquisition.
Utilizing knowledge context in virtual collaborative work
2005, Decision Support Systems
The understanding of knowledge can be impaired if it is isolated from the proper context. Despite the importance of contextual information, there has been limited support for utilizing context in current knowledge management and collaborative systems. This paper presents a knowledge context model, called KC-V, which facilitates the use of contextual information in virtual collaborative work. Four benefits of using KC-V are suggested: evolutionary accumulation of knowledge aligned with collaborative activities, supporting the virtual team lifecycle, improved understanding by rich navigation paths, and searching for knowledge with similar context. A web-based collaboration system called VWSS is developed using KC-V.
The effects of contextualized access to knowledge on judgement
2001, International Journal of Human Computer Studies
This research conceptualizes contextualized access to knowledge, i.e. the ability to access task domain knowledge within the context of problem-solving and investigates its effects on knowledge dissemination. Two informationally equivalent versions of a financial analysis knowledge-based system (KBS) were compared in a laboratory experiment, one with contextualized access to the underlying task domain knowledge (deep explanations) via hypertext-style links and the other without such access. Results indicate that contextualized access had significant advantages. It afforded a major portion of the requests for deep explanations to occur in the context of problem-solving, as opposed to in the abstract, and led to a significant increase in the number of requests. The increased utilization of deep explanations and contextualized use were associated with a greater degree of congruence between users' judgement and KBS. The conclusion is that availability of knowledge alone is not sufficient;contextualized accessibility is the key for knowledge dissemination and for influencing performance.
Cooperative problem solving and explanation
1995, Expert Systems With Applications
Recent studies have pointed out several limitations of expert systems regarding user needs, and have introduced the concepts of cooperation and joint cognitive systems. While research on explanation generation by expert systems has been widely developed, there has been little consideration of explanation in relation to cooperative systems. Our aim is to elaborate a conceptual framework for studying explanation in cooperation. This work relies heavily on the study of human-human cooperative dialogues. We present our results according to two dimensions; namely, the relation between explanation and problem solving, and the explanation process itself. Finally, we discuss the implications of these results for the design of cooperative systems.
Developing an explanation component for a knowledge-based system: Discussion
1995, Expert Systems With Applications
The articles in this special issue cover a wide range, and represent many different conceptions of “explanation” and what explanation research should be about. In this short discussion article, I will address some of the implications of this work for those trying to develop a practical explanation system. In particular I will focus on how explanation needs are determined, how detailed explanation content is chosen, and how the explanations provided relate to the problem solving activities of the system.

View all citing articles on Scopus

^∗: The authors are listed in alphabetical order. Dr. Mittal did the research and wrote the paper while doing his PhD at USC/ISI, in the Department of Computer Science at the University of Southern California. Dr. Paris is also a research professor in the Computer Science Department at the University of Southern California.

Copyright © 1995 Published by Elsevier Ltd.

Machine learning models to detect anxiety and depression through social media: A scoping review
Computer Methods and Programs in Biomedicine Update, Volume 2, 2022, Article 100066
Arfan Ahmed, …, Mowafa Househ
A method to detect and locate faulted area in distribution systems using the existing measurements structure
International Journal of Electrical Power & Energy Systems, Volume 131, 2021, Article 107073
João Tito Almeida Vianna, …, Débora Rosana Ribeiro Penido
AI and deep learning for processing the huge amount of patient-centric data that assist in clinical decisions
Cognitive and Soft Computing Techniques for the Analysis of Healthcare Data, 2022, pp. 101-121
João Alexandre Lôbo Marques, …, Simon James Fong
Impact of lockdown measures to combat Covid-19 on air quality over western Europe
Science of The Total Environment, Volume 741, 2020, Article 140426
Laurent Menut, …, Arineh Cholakian
Understanding consumer intentions toward cross-border m-commerce usage: A psychological distance and commitment-trust perspective
Electronic Commerce Research and Applications, Volume 39, 2020, Article 100920
Yi Cui, …, Kerry Kurcz
Exploring the relationship between entrepreneurial resilience and success: The moderating role of stakeholders' engagement
Journal of Business Research, Volume 119, 2020, pp. 142-150
Gabriele Santoro, …, Elena Candelo