Title:
Persuasion, Strategic News Sharing, and Cascades on Social Networks
Persuasion, Strategic News Sharing, and Cascades on Social Networks
Time:
02/12 (Sat.) 11 am PST, 12 pm MST, 1 pm CST, 2 pm EST, 8 pm CET
02/13 (Sun.) 3 am Taiwan
Time zone conversion tool
02/12 (Sat.) 11 am PST, 12 pm MST, 1 pm CST, 2 pm EST, 8 pm CET
02/13 (Sun.) 3 am Taiwan
Time zone conversion tool
Keywords:
Economics, Game Theory, Network Economics, Information Economics, Persuasion, Strategic News Sharing, Spread of Information, Social Networks
Economics, Game Theory, Network Economics, Information Economics, Persuasion, Strategic News Sharing, Spread of Information, Social Networks
Abstract:
We develop a game-theoretic model of sharing decisions among online users of a Twitter-like social network. Each agent has a subjective prior on an unobservable real-valued state. When receiving news, agents make a decision as to whether they should share the news with their followers based on how persuasive the news may be in moving their followers’ opinions closer to theirs, assuming a nominal cost for sharing. We characterize the dynamics of spread as an endogenous Susceptible-Infected (SI) epidemic process and derive agents’ sharing decisions and the size of the cascade spread at the equilibrium of the corresponding game. We show that low credibility news can result in a larger cascade than credible news when the network is highly connected. We further show that increased polarization in prior beliefs in the population prompts more sharing of lower credibility news, resulting in larger cascade size. Finally, we fully characterize the relationship between cascade size, network connectivity, and news credibility in terms of polarization and diversity of prior beliefs. Our results provide a theoretical foundation for recent empirical observations demonstrating faster and wider spread of low-credibility and false information on social networks. This is a joint work with Amir Ajorlou and my advisor Ali Jadbabaie.
We develop a game-theoretic model of sharing decisions among online users of a Twitter-like social network. Each agent has a subjective prior on an unobservable real-valued state. When receiving news, agents make a decision as to whether they should share the news with their followers based on how persuasive the news may be in moving their followers’ opinions closer to theirs, assuming a nominal cost for sharing. We characterize the dynamics of spread as an endogenous Susceptible-Infected (SI) epidemic process and derive agents’ sharing decisions and the size of the cascade spread at the equilibrium of the corresponding game. We show that low credibility news can result in a larger cascade than credible news when the network is highly connected. We further show that increased polarization in prior beliefs in the population prompts more sharing of lower credibility news, resulting in larger cascade size. Finally, we fully characterize the relationship between cascade size, network connectivity, and news credibility in terms of polarization and diversity of prior beliefs. Our results provide a theoretical foundation for recent empirical observations demonstrating faster and wider spread of low-credibility and false information on social networks. This is a joint work with Amir Ajorlou and my advisor Ali Jadbabaie.
No comments:
Post a Comment