FActCheck: Keeping Activation of Fake News at Check
FActCheck: Keeping Activation of Fake News at Check
复制标题
DOI：
--
复制
发表时间：
2018-07
期刊：
影响因子：
--
通讯作者：
Ajitesh Srivastava;R. Kannan;C. Chelmis;V. Prasanna
中科院分区：
文献类型：
--
作者：
Ajitesh Srivastava;R. Kannan;C. Chelmis;V. Prasanna
近年来，假新闻的传播已成为一个关键问题。与之斗争的一种方法是传播相应的真实的新闻。为了达到这个目的，我们找到一组有可能接收到假新闻的个体，这样他们就可以检验它的可信度，当他们传播相应的真实的新闻时，很可能会到达大量的个体。对于这个问题，我们提出了一个多项式时间贪婪算法（AFC），提供（1-1/e-e）-近似。我们通过开发一种快速图修剪启发式算法（RAFC）来进一步优化AFC的运行时间，该算法在检查假新闻的传播方面表现良好。我们在真实网络上的实验表明，我们的方法优于社会网络分析文献中流行的方法。
The diffusion of fake news has become a crucial problem in recent years. One way to battle it is to propagate the corresponding real news. To achieve this goal, we find a set of individuals who are likely to receive the fake news so that they can test its credibility, and when they propagate the corresponding real news, it is likely to reach a large number of individuals. For this problem, we propose a polynomial time greedy algorithm (AFC) which provides (1-1/e-e)-approximation. We further optimize the runtime of AFC by developing a fast graph-pruning heuristic (RAFC) that performs as well as AFC in checking the spread of fake news. Our experiments on real-world networks demonstrate that our approach outperforms popular methods in social network analysis literature.