A Semantic-Based Friend Recommendation System for Large-scale System

Abstract

Informal community locales have pulled in a large number of clients with the social revolution in Web 2.0 . Most informal organization sites depend on individuals' vicinity on the social diagram for friends suggestion. Existing work have a tendency to present Match Maker, a cooperative filtering friend recommendation system supported temperament matching. The goal of Match Maker is to leverage the social data and mutual affection among individuals in existing social network connections,and turn out friend recommendations supported made discourse information from people’s physical world interactions. Matcher permits users’ network to match them with similar TV characters, and uses relationships within the TV programs as parallel comparison matrix to recommend to the users friends that are voted to suit their temperament the most effective. The system’s ranking schema permits progressive improvement on the temperament matching, accord and a lot of various branching of users’ social network connections. Lastly, our user study shows that the appliance can even induce a lot of TV content consumption by driving users’ curiosity within the ranking method.