Privacy-Preserving Link Prediction Method for Personalized Recommendation in Social-Attribute Network

Résumé fourni par la source

Personalized recommendation services have recently been widely provided in online social networks (OSNs). OSNs have large number of users, and users’ information and browsing data are saved in online services due to frequent communication between users. To achievebetter performance, personalized recommendation services need users’ characteristics and behavior information, which brings privacy issues into concern. Therefore, balancing privacy preservation and recommendation results has become the main focus in this field. In this paper, we combine a privacy preserving method with the social network model and make full use of the user’s attribute information in the social network to improve personalized recommendation results based on the privacy-preserving link prediction(PPLP) framework. Additionally, a link prediction algorithm with attribute classification is proposed in this paper, which considers the connections between user attributes and the similarity between users. The improved PPLP was evaluated on Google+ datasets and the results show that it can improve the accuracy of recommendation results while protecting users’ information.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.