News Feeds Optimization via Users´ Interaction Graph over Mobile Social Networks

Online social networks are increasingly playing roles of propagating news. We are surrounded with diverse data every day in today´s information exploding era and suffer from the littery contents. Furthermore, due to most users of online social networks have turned their platforms from PCs to smart phones, it is a challenge to select contents that users interested in with resource constraints in mobile social networks. Considering the contradiction between large volume of data and memory constraints in mobile social networks, we present an efficient news filtering mechanism via users´ interaction to avoid unnecessary display of data that barely read. By leveraging the Bloom filter technology, we can record users´ interaction dynamically with a slight false positive, and filter the little possible readable updates of friends from their News Feed sites. We conduct a simulation to evaluate our design. Results show significant performance for storage saving and high filtering rate.