![]() |
|||
![]()
|
![]() |
![]() |
![]() |
2.2. USABILITY ISSUES WITH EXPERT SYSTEMSExpert systems raise new design issues concerning three separate but highly interacting aspects of a design process: knowledge capture, encoding the knowledge and expertise into the system, and ensuring that the system is accepted by an active user community. The designer of a user interface must consider many aspects of computer usage ranging from cognitive models of users to ergonomic issues. The designer must concentrate on many aspects of usability, including a focus on users and their tasks, getting empirical evidence about effectiveness, and stressing iteration between designers, implementers and users. It was sugggested in the introduction chapter of Hendler's book (1988) that these three aspects are special for user interfaces for expert systems:
2.3. SPECIAL NEEDS FOR EXPERT SYSTEM INTERFACESExpert systems have some special interface needs. One of the most essential requirements is the need for explanation of the reasoning process. The requirement becomes crucial to the acceptance of the system by users. An expert system is not just a tool that implements a process; rather, it is a representation of that process. Many of these processes correspond to judgments that can have critical consequences in the real world. The user interface must present not only final conclusions and recommendations, but an explication of the processes of how such conclusions are reached. Expert systems are often used as a decision-making support tool. The interface designer must be able to provide support for a knowledge engineer who is trying to enter and encode the representation of the process as well as users who wish to see the process of decision-making per se. 2.4. USER MODELING AND ADAPTIVE USER INTERFACESUser-centered system design in HCI emphasizes the importance of addressing users and their needs. On the other hand, researchers and practitioners have realized that the needs of users are not a constant, which implies that user interfaces must be adaptive so as to fit smoothly into the users' learning processes. Learner-centered design draws attention to the changing needs of users, both students and professionals, as they gain expertise and how these changes need to be reflected in the interface. Interface design must be tailored to support users as learners with case studies of their experiences in designing adaptive and adaptable interfaces for learners. User modeling has made considerable progress, particularly in the last few years. The need has been recognized in many application areas for software systems to automatically adapt to their current users. As a result, research in user modeling has extended into many disciplines that are concerned with the development of interactive computer systems used by diverse groups of users. These fields include Intelligent Interfaces, Active and Passive Help Systems, Hypertext Systems, Intelligent Information Retrieval, Natural-Language Systems, Intelligent Tutoring Systems, and Cooperative Expert Systems. The following is a taxonomy that has been commonly used to organize a broad range of classification parameters for adaptive user interfaces, including tasks and agents, types, levels, scope, goals, methods, strategies, models and architectures of adaptation, and adaptation techniques. The taxonomy is presented as a classification tree.
In addition, a number of formalisms and techniques are associated with user interface design, including user modelling, task modeling, dialog modeling, and plan recognition.
|
![]() |
|
Use of this site is subject certain Terms & Conditions. Copyright (c) 1996-1999 EarthWeb, Inc.. All rights reserved. Reproduction in whole or in part in any form or medium without express written permission of EarthWeb is prohibited. Please read our privacy policy for details. |