A New Method for Identifying Users Interest for Personalized Recommendations

Social Media encourages users to participate more interactions in Internet. They could share, interact, post the activity. In social networks relation could be defined by post and like to each other status. This data is like a treasure vault waiting to be utilized by the system to develop the recommendation systems. We propose a novel method to make personalized recommendation system which utilizes the user affecting index, user interest, user influences and familiarity between users. There are three purposes in this paper. The first one is to find the central user of social network and his/her influences to the other users. The next one is to find the correlation between user's attribute and other user's in social network. Finally, to discover the opposite users those have the least influences. The relationships of users and users could be utilized to make recommendation of items in social media.