A decentralized AI system for transcript applications

A working expert system for assessing a student's course requirements for graduation in various departments of a university is presented. The system draws on artificial intelligence (AI) techniques of expert systems and natural language processing to facilitate interaction with both intermediate- and end-user. Expertise is expressed in rules that take the form of English sentences. The flexibility of the AI approach allowed substantial benefit from an earlier related system for admission of transfer students. A key feature with respect to the workplace is the system's transparency or non-disruptiveness, achieved by using an existing database. A special editor for output that looks like a filled-in form of the type that has traditionally been prepared by hand is provided. Transparency and ease of use have enhanced acceptability of the system for the people in the registrar's office, who have enthusiastically collaborated in design and testing. Acceptability should also make decentralization to the departments, the university's counterpart to the government field office, possible.<
>