Enhancing fuzzy inference system based criterion-referenced assessment with an analogical reasoning schema

In this paper, a fuzzy inference system (FIS) that incorporated with an analogical reasoning schema based criterion-referenced assessment (CRA) is proposed. The aim of CRA is to report students’ achievement with reference to a set of objective reference points. Usually, scores were given to each task in order to eases the assessment as in common practice. A total-score is further obtained with a simple addition or weighted addition of these scores. Scoring rubric is an essential tool for subjectivity assessment. A search in literature reveals that the use of FIS in CRA is not new. It can be explained as an alternative approach how a total-score can be obtained. For a multiple input FIS based CRA, a large set of fuzzy rules are required. With the use of grid partition, the number of fuzzy rules required increases in an exponential manner and this phenomenon is known as the curse of dimensionality or combinatorial rule explosion problem. It is a tedious work in getting a full set of rules. The main objective of this paper is to propose a novel FIS based CRA schema that allow rules to be reduced. We suggest to adopt a systematic approach to select a set of rules (from the full rule base), and to incorporate an analogical reasoning schema to predict unknown consequent. An FIS based CRA procedure with an analogical reasoning schema is proposed and evaluated with a case study relating to students’ laboratory project assessment is conducted in UNIMAS.