Improving e-Commerce Severity Rating Measurement Using Consistent Fuzzy Preference Relation

Abstract

Usability is a critical success factor in business. However, many usability issues on e-commerce websites cannot be adequately handled because there is no priority scale. The severity rating helps focus on the significant problem. The severity triggered by the data is considered better than the severity triggered by the evaluator. Data obtained quantitatively used to determine essential values and usability scores. Previous research has applied the Fuzzy Analytical Hierarchy Process (FAHP) method with Extent Analysis (EA), and Fuzzy Preference Programming (FPP) approaches. We were assessing the weight of criteria to evaluate usability and determine severity rating. However, EA and FPP approaches have disadvantages. Among them, when determining the number of paired comparisons at the level of importance between criteria. The number of comparisons that must be assessed by Decision-Maker (DM) causes the fuzzy pairing matrix to be inconsistent. This inconsistency causes the weight between rules to be invalid. The Consistent Fuzzy Preference Relation (CFPR) method is present to overcome the problem of the number of paired comparisons. The CFPR method summarizes the comparison steps to facilitate DM in assessing the level of importance between criteria. The results show the results of rank similarity testing; the EA and CFPR methods have close relationships. The FPP and CFPR methods have a weak correlation in generating usability ranking and severity ratings.