Detecting social bots activity in Twitter using deep learning approach

Abstract

With millions of users, One of the most popular microblogging social networking sites is Twitter. Social bots (also known as spambots) are content-generating robots that exist on social media networks. This study provides an overview of new strategies for distinguishing between social bot accounts and human accounts that have recently appeared. Users have benefited from the deployment of social bots to automate analytical services and improve service quality. On the other hand, social bots have been employed to distribute false information, which can have real-world consequences. The describe a method for determining if that user is a Bot or a human. The propose recognizing such accounts by evaluating message predictability because communications published from the same bot account usually follow basic, repeating patterns. The shallow and deep learning algorithms for tweet-based bot detection, as well as their performance outcomes, are also described. To determine if tweets were made by actual people or bots, a machine learning architecture is proposed. Customers actual voices can be greatly distorted by malicious bots. We categorized millions of open-source functions in the cresci 2017 datasets. The goal of this study is to look at social bots in the context of user-generated content deep learning.