A Fuzzy Approach for Personalized Product Clustering with Flexible Discriminating Power

This paper proposes an improved linguistic quantifier that operates with penalty function so that a set of products can be clustered into hierarchical levels. The method is based on fuzzy approach with an aim at personalizing the display of products in an order of personal preference on a set of product attributes. The discriminating power of the system can be flexibly tuned via the set up of quantifier´s parameters and the penalty function. Numerical example is given for illustrating the computation, providing comparative results to the existing method and giving some technical insights.