Applying Fuzzy Logic to Recommend Consumer Electronics

Abstract

Depending on the type of the product, different kinds of personalized recommender systems can be built to guide the consumers in a large product feature space. In the approach, we present a fuzzy-based recommender system for those products that a general consumer does not buy very often, especially for consumer electronic products. For those consumer electronic products, it is difficult and not necessary to reason a customer’s previous preferences because there may not be enough information about the customer’s past purchases and the customer may have his specific requirements in each single purchase. Hence the system has specific domain knowledge and capability to interact with the consumer. Experimental results show the promise of our systems.