Detecting Bot-Controlled Accounts on Social Media Using Deep Learning

Abstract

The proliferation of bot-controlled accounts on social media platforms poses significant risks to user trust and platform integrity. Recent approaches to bot identification suffer from imbalanced data, as well as overfitting and scalability issues. To address these challenges, this paper proposes a deep learning-based framework to detect such accounts using behavioral and content-based features extracted from Twitter data. The methodology integrates feature engineering, data preprocessing, and deep learning models. Evaluated on the Cresci-2017 dataset, the best-performing model achieved a test accuracy of 98.51%, with a precision of 99.16%, recall of 98.23%, and an F1-score of 98.69%. The results show that deep learning can effectively differentiate between genuine and bot-controlled accounts, contributing to enhanced security and authenticity in online interactions.