KESS: knowledge engineering support system

Building an expert system in medicine is a difficult, time-consuming job. It involves many meetings of people from various disciplines. In each of the sessions, statements about the relationships used in decision-making are made by experts, data are examined from medical literature and hypothetical relationships are tested against local patient data bases. In order to maintain continuity and provide future reference to this process, a relational information management system know as Knowledge Engineering Support System or KESS was created at the University of Utah. KESS utilizes many-to-many links with multiple files and central link files to track the different kinds of information generated and used in the four knowledge building projects. It is used on-line during the expert meetings or knowledge engineering sessions to enter new data or to revise and update information already entered in previous sessions. The data can be retrieved using indexed fields or by using macros to generate reports. The reports can be printed or sent via modem to the expert at his or her workstation. KESS uses a commercial software that has the advantage of being immediately available, thereby avoiding the design of a database management systems from scratch. It also has the advantage of allowing easy and safe data transfer between different version of KESS program. The condition of the development and the use of KESS will be described in more detail in the thesis.