On Flow Authority Discovery in Social Networks.

摘要: A central characteristic of social networks is that it facilitates rapid dissemination information between large groups individuals. This paper will examine the problem determination flow representatives, a small group authoritative representatives to whom piece leads maximum spread. Clearly, affected by number different structural factors such as node degree, connectivity, intensity interaction and global behavior underlying network. We propose stochastic model, use determine in first design an accurate RankedReplace algorithm, then Bayes probabilistic model order approximate effectiveness this algorithm with fast algorithm. results on real network data sets, show method more effective than state-of-the-art methods.