This paper presents a contribution to design an online preference based system. The objective of the system is to assist a customer in the products selection process. The product considered here to describe the application is the comic. Current e-commerce recommendation systems assist customers in this process. Nevertheless, quality of the recommendations produced remains a real challenge. There are products that are not recommended to customers though they would appreciate them and others recommended to them though they do not appreciate them. Quality and relevance of recommendations is addressed in this paper. The customer's choice is not only based on product characteristics but also on his/her perceptive expectations. Therefore, products recommendations are considered relevant since they meet customer's expectations and particularly perceptive ones. The suggested algorithm aims to recommend spontaneously comics to an active customer. It is mainly based on collaborative filtering and neighbourhood formation. Cluster of neighbours is formed. Neighbours share common perceptive preferences with the active customer. Favourite comics of the neighbours may interest the latter. Selection of the appropriate comics is based on the product characteristics expected by the active customer. The algorithm produces relevant comics which meet the active customer's expectations and particularly perceptive ones.