基于部分路径的社交网络信息源定位方法

Source Localization Algorithm Based on Partial Paths for Social Networks

随着微博、微信等在线社交网络的快速发展,社交网络上的不实信息呈现爆发式的传播,往往会引起严重的后果,如何寻找谣言等不实信息在社交网络中的传播源头具有重要的应用意义。该文提出一种面向在线社交网络的信息源点定位方法,与现有的基于观察点的定位方法不同,该方法考虑了传播过程中信息普遍带有的部分传播路径,并以此重构传播过程,修正传播子图,从而更准确地定位信息源点。在模型网络及实际网络上进行实验,说明了该算法的有效性。

With the rapidly growth of online social networks such as microblog and WeChat, the false information breaks out on the social network and often brings serious consequences. How to locate the rumor source is of great importance for many applications. This paper proposes a source localization algorithm on online social network. We consider the characteristic that the information often contains some partial spreading, and design a more accurate algorithm to locate the information source. The results show that the improved algorithm can provide a more accurate spreading trees and improve the localization accuracy. Experiments on model and real network show the effectiveness of the improved algorithm.