Degree accreditation report auto-generation by logic encoding and processing

Abstract

Maintaining the accreditation profile of an academic programme is a key activity in so-called professional degrees such engineering, commerce and law. The complexity of the accrediting criteria tends to rise over time as accrediting bodies require quantitative evidence of competence of increasingly specific graduate attributes. Evaluation of graduate attributes may therefore require complex logic processing which challenges the human capacity. This has the negative side effect of discouraging curriculum revision not for pedagogic reasons but simply due to the complexity of evaluating complex logic patterns against a data set whose structure is shifting. These challenges can be overcome through the application of logic encoding and processing. A computing system is better suited to such processing tasks since logic processing is fundamental and well-established to such systems. On the other hand, the efficient representation of a complex accreditation logic rule base then becomes the challenge. This paper describes the representation of the accreditation logic of eight engineering academic programmes at the Durban University of Technology through the AutoScholar Advisor System in preparation for evaluation by the Engineering Council of South Africa. It is shown that the system generates accurate reports even with deeply nested logic structures and with changes in curriculum over time.