Journals Proceedings

Abstract

The users of social network sites are always constructing a large network, recommending friends to registered users is a crucial task for these sites. Traditional content- based or collaborative filtering recommend technologies are limited for the task, because the users and items are the same dataset, which have rich attributes and complex social graph. In the paper, we proposed a local random walk and random forests combined method to do friends recommendation for large social networks. The experimental results show that the method has high precision and high recall at the same time.