Bilkent University
Department of Computer Engineering


Modeling and Mitigating Online Misinformation Propagation using Game Theory and Blockchain Crowdsourcing


Tolga Yılmaz
PhD Student
Computer Engineering Department
Bilkent University

Online misinformation has been a source of disruption for social media users, leading them into believing falsely served information on crucial issues. In this work, we first investigate the similarity of misinformation propagation in online social networks to evolutionary games on graphs. Our simulations show that misinformation propagation can be modeled as a game and the spread of misinformation is at an advantage over correct information in online social networks. Although there are efforts to detect misinformation using automated machine learning based systems or by manual categorization by human annotators, we argue that in order for them to be reliable, they need to be decentralized. For that purpose, we propose a decentralized blockchain based system that utilizes crowdsourcing to validate information. The system employs a reputation model which is used to determine user influence. According to our simulations, the system shows promising results for troll resistance.


DATE: 26 March, 2018, Monday, CS590 & CS690 presentations begin at @ 15:40