Friend book: A Scalable and Efficient Friend Recommendation Using Integrated Feedback Approach

Abstract

Friend book is a novel semantic-based friend recommendation system for social networks, based on their life styles instead of social graphs which recommends friends to users. Friend book discovers life styles of users, measures the similarity of life styles between users, if their life styles have high similarity it recommends friends to users. User’s daily life is modelled as life documents, from which users life styles are extracted by using the Latent Dirichlet Allocation algorithm; Similarity metric to measure the similarity of life styles between users, user’s impact is calculated in terms of life styles with a friend-matching graph. A linear feedback mechanism is integrated that exploits the user’s feedback to improve recommendation accuracy. In this paper various recommender systems are classified are discussed. This paper focuses on providing the overview about the various categories of recommendation techniques developed till now. This paper we present review on recommendation system for find friend on social networks.

Key words: Recommendation system; social networks; friend; item; user