When a user interested in a service/item, visits an online web-portal, it provides description of its interest through initial search keywords. The system recommends items based on these keywords. The user is satisfied if it finds the item of its choice and the system benefits, otherwise the user explores an item from the list. Usually when the user explores an item, it picks an item that is nearest to its interest from the list. While the user explores an item, the system recommends new list of items. This continues till either the user finds its interest or quits. In all, the user provides ample chances and feedback for the system to learn its interest. The aim of this paper is to exploit the user-generated responses in the same session. One can further utilize the history (e.g., previous user ratings) to design good recommendation policies. We develop algorithms that efficiently utilize user responses to recommended items and find the item of user's interest quickly. We first derive optimal policies in the continuous Euclidean space and adapt the same to the space of discrete items. In the continuous Euclidean space, the optimal recommendations (e.g., with two recommendations) at the same time step are at 180 degrees from each other, while are at 90 degrees with respect to the ones at the previous time step. We propose the notion of local angle in the space of discrete items and develop user response-local angle (UR-LA) based recommendation policies. We compared the performance of UR-LA with widely used collaborative filtering (CF) based policies on two real datasets and showed that UR-LA performs better in majority of the test cases. We also proposed a hybrid scheme that combines the best features of both UR-LA and CF (and history) based policies, which outperforms them in most of the cases.