The goal of service selection issue is to generate suitable service to the customer. The selection was addressed by many approaches to face into the riches and the variety of services available online. Currently the users cannot find what they want easily, so to overcome this challenge a recommender system is needed. This system is widely applied in many fields such as commercial sites, social networks and service oriented architecture. It allows guiding of the user in their navigation, based on their previous choices or on its neighbors. Despite the advantages of such systems there are always limitations. We aim through this paper to tilt the angle of recommendation by intermixing between user intention and recommendation, to improve the selection issue.