In the context of electronic commerce, recommender systems enable merchants to assist customers in finding available products that will best satisfy their need. However, a recommender system usually operates as a kind of black box from which customers receive recommendations for products. Of particular interest are recommender systems based on Collaborative Filtering, in which customers provide the recommender system with ratings on products and receive recommendations based on the similarity paradigm. In this paper, we introduce an approach in which customers self-control the collaborative filtering process. More precisely, our approach uses a list of contacts, which allows a better control on the recommendations provided locally (using the list of contacts), while still being able to access and even influence the global recommendations of the system (using the whole database of customers). We believe that our system allows more confidence in the customer because it enables him to maintain real-time statistics on his similarity with customers forming his ring of contacts.