Discovering and drawing out the relationship between users and items in a service-based companies or organizations are the essence of a recommendation system. Here we address a novel approach for the recommendation system, incorporating the means of collaborative aspect between the users internal hidden patterns and the items or goods to be recommended. Unlike the existing methods, our algorithm introduces a guiding factor between the user hidden state and the choice over the item set, such that it gives additional degree of freedom for the recommendation system to opt on which factor is more prominent.