A dynamic Web Recommender System using Hard and Fuzzy K-modes clustering

Abstract

This paper describes the design and implementation of a new dynamic Web Recommender System using Hard and Fuzzy K-modes clustering. The system provides recommendations based on user preferences that change in real time taking also into account previous searching and behavior. The recommendation engine is enhanced by the utilization of static preferences which are declared by the user when registering into the system. The proposed system has been validated on a movie dataset and the results indicate successful performance as the system delivers recommended items that are closely related to user interests and preferences.