Abstract
We investigate the problem of source detection in information spreading throughout a densely-connected network. Previous works have been developed mostly for tree networks or applied the tree-network results to non-tree networks assuming that the infection occurs in the breadth first manner. However, these approaches result in low detection performance in densely-connected networks, since there is a substantial number of nodes that are infected through the non-shortest path. In this work, we take a two-step approach to the source detection problem in densely-connected networks. By introducing the concept of detour nodes, we first sample trees that the infection process likely follows and effectively compare the probability of the sampled trees. Our solution has low complexity of O ( n 2 log n ) , where n denotes the number of infected nodes, and thus can be applied to large-scale networks. Through extensive simulations including practical networks of the Internet autonomous system and power grid, we evaluate our solution in comparison with two well-known previous schemes and show that it achieves the best performance in densely-connected networks.
Highlights
During the last decade, mobile Internet services have become very popular
We investigated the problem of source detection in general networks, in particular in densely-connected networks, and develop low-complexity approximations with high detection performance
Incorporating our intuition, we developed a new evaluation scheme, the Growth-Tree Estimation (GTE) algorithm, that approximately compares the probabilities by taking into account all the edges connected to VI
Summary
Mobile Internet services have become very popular. In particular, Social Networking Services (SNS) like Facebook and Twitter have emerged and attracted several billions of users. There have been several studies on the information source detection in the literature [1,2,3,4,5,6,7,8,9], most of them have assumed a specific type of network topology, i.e., trees, and cannot be directly applied to a more general network topology that most social networking services have Their extension to general networks is not straightforward and often suffers from high computational complexity. In a general network graph, there are multiple possible paths of the infection due to the loops or cycles in the underlying topology To this end, we generate a spanning tree that consists of the edges that likely infect a node, by introducing the concept of detour nodes.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
