Knowledge that Transforms

Dialogue management involves the application of systems in the design and delivery of user-system dialogues. This article describes a dialogue management environment that is based on a concept of dialogue independence: the separation of dialogue definition from dialogue execution. In this environment, software facilties provide building blocks for design and delivery of dialogues that are consistent, efficient, shareable, and portable. The environment supports multiple workstation and communication protocol environments. This environment is compared to those emerging in homogeneous hardware environments such as HyperCard.

Organizations are undergoing radical changes in both their use of technology and their basic practices. We can expect that these changes will accelerate as the pressures continue to grow. Managers are faced with radical restructuring initiatives to support the downsizing, downscaling, and delayering of objectives. The growth of interfunctional teams and often cross-organizational teams is leading to further initiatives in the establishment of and cooperative clusters of both short and long term duration. Integration within and across the organizational boundaries i further stimulating interest in leveraging information technologies to enable and support work of groups and teams. Whether these are teams with a specific mission, standing committees that have a regular or recurring work schedule, or specially assembled groups that will have little cooperation beyond the current task at hand, each has different interests in applying information technologies to support these meetings and other group work.

Since models play a critical role in human decision processes, model management is considered a very important function for decision support. This article examines how model management systems can be designed to support group problem-solving. First, basic concepts of model management and functional requirements for group model management systems are described. Then, an architecture for group model management systems design is presented. Finally, major implementation issues are discussed.

A joint venture between a university-based research institute and a health insurance company to build a knowledge-based system to perform medical review of health insurance claims is described. In an early formal test, the system made appropriate decisions for approximately 70% of the claims. The remaining claims were referred to human reviewers in accordance with company policy. Forming a joint venture proved to be a reasonable cost alternative to distant and more expensive consultants. The article examines the impact of the differing cultures of the company and the university on the cohesion in the joint knowledge engineering group. It also examines the current literature on the development cycle for building knowledge-based systems as the framework for analyzing the events in this project, particularly the influence of the claims review task on system design. From another perspective, it examines participant roles in terms of shifts of attention among domain knowledge, knowledge representation, system performance, and the kinds of skills needed to improve the evolving system. The conclusion includes a series of recommendations that may assist other companies and universities setting up similar joint ventures.

Many years ago, MIT's Joe Weizenbaum, frightened by the simple-mindedness with which his Eliza experiments had been misunderstood, expressed serious reservations about artificial intelligence's implicit assumptions and directions. He came to be seen as an adversary of those who had been colleagues, as a whistle-blower, and that particularly dangerous opponent, a convert. Whether or not he intended to, he went out on an academic, intellectual, even spiritual limb. His work has had significant effect on people outside the computing-related disciplines, but is for the most part rejected or ignored by those inside them. The adversarial approach to the social implications of I.T. is not to be recommended, because of the intellectual isolationism it can lead to, and its inherently limited effectiveness.

Software maintenance continues to be one of the most complex problems facing much of the software industry. This study examines EDP professionals’ perceptions of what makes software maintainable. Perceptions were assessed through a series of factor analyses on the data collected from a survey of 149 professionals from 20 organizations of varying sizes. The respondents were asked to rate the relative importance of various features which contribute to the ease of maintenance. Results showed that EDP professionals do not share a common perception of software maintainability: software project managers have different perceptions of what contributes to the ease of maintenance than those who are directly involved in maintenance operations. If those who manage and those who actually maintain software do not, in fact, agree on the importance of maintenance factors, then the maintenance problem is much more complex than EDP professionals or software engineers might have originally thought. Perhaps this is the reason why, according to the survey, EDP professionals are less than optimistic in general about the ability of state-of-the-art approaches to solve the problems of software maintenance.

This paper provides a new approach for managing the design and delivery of information and decision support systems for strategic decision making. It draws on experiences gained from implementing systems and services for enhancing the strategic decision-making process of the Cabinet of Egypt. The article challenges the conventional views of conceptualizing decision support systems and methods for managing them. It introduces an “issue-based” management method for the design and delivery processes. The distinctive features of this approach include a focus on issues rather than decisions, a distinction between information support services and decision support services, prototyping the management of delivery as well as design, and a dynamic tracking back-end. Finally, the article compares the conventional and issue-based DSS approaches. Such a comparison suggests that the issue-based approach can be an effective stepping stone for the design and delivery of executive information systems (EIS) in corporate contexts by providing DSS that are “EIS- ready.”

This article proposes a comprehensive framework for comparing the current methodologies and tools for information engineering and using these methods for applying information technology to construct the overall information systems architecture for the organization. The proposed framework consists of two dimensions: (1) an expanded sequence of the traditional system life cycle; and (2) the conceptual depth of the methods. The article shows that information engineering is the key to effective information management. Using the proposed framework, 26 widely-cited methods for information engineering are compared. Evolution to more effective methods of information engineering are needed to align future information systems requirements to strategic goals and objectives of an organization and to exploit the current information systems technologies for competitive advantage.

Traditional cost-benefit approaches to evaluating the effectiveness of decision support systems are generally regarded as inadequate. More recent approaches account for the intangible benefits overlooked by traditional costbenefit analysis, but still attempt to express benefits in terms of costs. This paper focuses on a value analysis. It advocates a methodology for (1) identifying significant intangible benefits for a specific decision support system, (2) quantifying these benefits in value terms, and (3) establishing a decision rule for identifying the significance of the proposed system. Finally, an alternative approach at deriving the benefit utilities is examined.