Digital Equipment Corporation -
O U R GOAL IS TO MAKE PRO-gramming easier, SO that creating useful programs takes less time and effort, and little or no programming skill. Many work environments do not receive enough help from application programs because creating such programs requires time and highly specialized skills, both of which are costly. The programmer must learn a great deal about how users perform tasks; this communication process is slow and susceptible to errors. Also, mastering programming languages and CASE tools requires extensive training. Empowering people to build their own programs might reduce these difficulties.The main problems in building application programs arise from the nature of real tasks. The full description of a task is messy and filled with details that make the task unique. This uniqueness is important to the people involved in the taskit's how they differentiate the task from other similar tasksbut the messiness makes it difficult to describe an abstract, computational process for performing the task (a sidebar discusses similar approaches to this problem). This messiness also prevents the early versions of a program from covering the task well: Tasks change continuously as the user population, the kinds of infor-
KNACK is a specialized knowledge-acquisition tool that generates expert systems for reporting tasks. The tool derives its power from exploiting the presupposed acquire-and-present problem-solving method used by the expert systems it generates. The method incrementally acquires relevant information and produces a report. It can also be combined with other problem-solving methods. An important goal in the development of KNACK is to create a tool that elicits knowledge from domain experts without requiring knowledge-engineering skills on their part. To reach that goal, KNACK's approach to knowledge acquisition uses existing AI techniques to derive a general description of how to acquire and present information from a specific sample description.
IntroductionExisting expert systems have proved that AI techniques can be used to solve a variety of knowledge-intensive problems. But expert systems are timeconsuming to develop and difficult to maintain. A key issue in developing any expert system is how to update its large and growing knowledge base. It has been shown that a large knowledge base can be kept maintainable by organizing it according to the roles that knowledge plays [Chandrasekaran 83, Clancey 83, Neches 84]. Based on this realization, a variety of knowledge-acquisition tools have been produced during the past years to overcome development and maintenance problems.Existing knowledge-acquisition tools focus on different aspects of the knowledge-engineering task. For example, KREME
In the past decade, expert systems have been applied to a wide variety of application tasks. A central problem of expert system development and maintenance is the demand placed on knowledge engineers and domain experts. A commonly proposed solution is knowledge-acquisition tools. This paper reviews a class of knowledge-acquisition tools that presuppose the problem-solving method, as well as the structure of the knowledge base. These explicit problem-solving models are exploited by the tools during knowledge-acquisition, knowledge generalization, error checking and code generation.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.