Mitigating Misinformation in Online Social Network with Top-k Debunkers and Evolving User Opinions

Online social networks provide an easy platform to share the information, and the spread of fake news and rumors has become prevalent, with severe consequences on major events including the US and Jakarta elections. Existing works have designed methods to find a set of top-k users to launch truth-campaigns and mitigate the negative influence of misinformation. The assumption is that these top-k users are open and willing to disseminate fact-checked content. Further, these methods assume that as misinformation and counter messages propagate in the network, user opinions once formed do not change. In this work, we address a more realistic scenario where users’ opinion can fluctuate before some deadline, and the goal is to find a good seed set of users from a set of debunkers to minimize the impact of misinformation. We propose a new opinion model that takes into account users’ biases and their social neighbours’ opinions. Based on this model, we design a mitigation solution to identify a subset of debunkers that maximizes the number of users who have been exposed to the misinformation but choose to believe the counter message. Experiments on Facebook and Twitter datasets demonstrate that our proposed solution is effective in mitigating the negative influences of misinformation.