Adversarial machine learning for protecting against online manipulation

TL;DR

This paper discusses how adversarial examples, typically used to attack machine learning systems, can also be exploited to improve model robustness against online manipulation, especially in fake news and social bot detection.

Contribution

It provides an overview of leveraging adversarial examples as tools to enhance the resilience of models against manipulation in online environments.

Findings

Adversarial examples can be used to strengthen models against attacks.

Improved detection of fake news and social bots using adversarial techniques.

Frameworks for applying adversarial training in online manipulation detection.

Abstract

Adversarial examples are inputs to a machine learning system that result in an incorrect output from that system. Attacks launched through this type of input can cause severe consequences: for example, in the field of image recognition, a stop signal can be misclassified as a speed limit indication.However, adversarial examples also represent the fuel for a flurry of research directions in different domains and applications. Here, we give an overview of how they can be profitably exploited as powerful tools to build stronger learning models, capable of better-withstanding attacks, for two crucial tasks: fake news and social bot detection.