Adding ReputationRank to member promotion using skyline operator in social networks

Abstract

Background: To identify potential stars in social networks, the idea of combining member promotion with skyline operator attracts people’s attention. Some algorithms have been proposed to deal with this problem so far, such as skyline boundary algorithms in unequal-weighted social networks. Methods: We propose an improved member promotion algorithm by presenting ReputationRank based on eigenvectors as well as Influence and Activeness and introduce the concept of skyline distance. Furthermore, we perform skyline operator over non-skyline set and choose the infra-skyline as our candidate set. The added ReputationRank helps a lot to describe the importance of a member while the skyline distance assists us to obtain the necessary condition for not being dominated so that some meaningless plans can be pruned. Results: Experiments on the DBLP and WikiVote datasets verify the effectiveness and efficiency of our proposed algorithm. Conclusions: Treating the infra-skyline set as candidate set reduces the number of candidates. The pruning strategies based on dominance and promotion cost decrease the searching space.