Walk globally, act locally: efficient influential user identification in mobile social networks.

Maotian Zhang,Panlong Yang,Chang Tian,Chaocan Xiang,Yan Xiong
DOI: https://doi.org/10.1145/2633675.2633680
2014-01-01
Abstract:ABSTRACTBeing a fundamental and challenging research topic, influential user identification has attracted much attention with the rapid growth of mobile social networks. Previous studies either focus on relatively stable network structure, or need fairly large information overhead in achieving global maxima. In tackling the dynamic topologies, we propose an influential user identification scheme fully exploiting the active mobile users, where the stable state property is leveraged under information potential construction scheme. We present an efficient routing scheme in reaching the global maxima without relying on specific routing protocols. We validate our scheme with both synthetic and real-world mobility traces. The experimental results show that, the proposed scheme achieves considerable performance on influential user identification and route construction, while bringing forth less overhead.
What problem does this paper attempt to address?