Election Manipulation on Social Networks with Abstention.

The Election Manipulation through the diffusion of (fake) news on social networks has been a subject that recently attracted the interest of many works in both in the communities of AI and social choice theory. However, all these works assume that each voter has to express her vote, not considering the possibility that she could abstain. One of the reasons of this omission is the lack of a satisfying modeling of how people choose to abstain. In this work, we try to fill this gap by presenting an innovative model for abstention that will match most of the real-world observations about the topic. Next, we will provide experimental evidence that abstention opportunity will help the manipulator to control the elections, by comparing how well-known algorithms and heuristics behave in the setting with and without abstention.