Social Media and Misleading Information in Democracy Using Machine Learning

Abstract

In this project, we address the challenge of mitigating the spread of misleading information in a democracy by designing a resource allocation mechanism tailored for strategic social media platforms. Recognizing the pivotal role of both a strategic government and the private knowledge of how misinformation impacts users, we propose a mechanism that incentivizes social media platforms to efficiently filter misleading content. The mechanism leverages an economically inspired approach to strongly implement all generalized Nash equilibrium, ensuring efficient filtering in the induced game. We demonstrate that the proposed mechanism is individually rational, budget-balanced, and guarantees the existence of at least one equilibrium. Furthermore, under quasi-concave utility functions and constraints, the mechanism admits a generalized Nash equilibrium and achieves Pareto efficiency. To validate our approach, we utilize machine learning algorithms—including Logistic Regression, Decision Tree Classifier, and Random Forest Classifier—to model and analyze the strategic behavior of social media platforms and the effectiveness of misinformation filtering.