With the rise of social network sites, almost every person has a profile. This profile contains various likes and dislikes about various items. It also contains their valence about other users on the network. In this paper a cloud-based recommendation engine is proposed which provides recommendations about various products given that a user likes a product X taking the balance theory of likes and dislikes over the network. This project can be used to provide personalized recommendations for advertisements or can be used by the user to explore other potential items that he may be interested in.