Recommender system based on user functional requirements using Euclidean fuzzy

Customer often feels difficult to describe what he needs when he intends to buy a high-tech product with complex features, such as smartphone, notebook, camera, PC, car, server, etc. It is because the most of users are less familiar with the technical features of this product types. However, most recommender systems had been developed still directly refer to product features when extracting user´s requirement. Naturally, users often express their needs based on functional requirements (need smartphone for gaming, reading e-book, online activity, etc). Considering the problems, this paper proposes a new approach for products recommendation based on functional requirements of product, to produce the more properly recommendation. Our proposed approach uses mapping between functional requirements, components product and supporting features of component. We utilize Euclidean fuzzy concept for calculating the similarity between user requirements and product features. The domain of our recommender system is smartphone area. The evaluation of the approach shows that the computation was accurate for 92.67%. The proposed approach also gives a more flexible recommendation process, since it considers the products that have better quality than the quality level that user wants.