Understanding consumer preference through fuzzy-based recommendation system

Abstract

We introduce a new recommendation technique based on fuzzy concepts such as the 2-tuple fuzzy linguistic representation and the fuzzy market research system approach. The recommendation technique uses general consumer perception about product attributes of multi-attributed products such as smartphones as input and gives consumers’ sequential preference of the products as output. The recommendation technique explains the power of an attribute in predicting other attributes, especially in fuzzy and uncertain environments, through the conceptualisation of consumers’ innate associations between the attributes. Our technique provides an understanding of consumer preference through the articulation of buyers’ trade-offs amongst attributes of a product.