Enhancing Social Media Integrity: A Multi-Platform Machine Learning Approach to Detecting Fake Accounts with a User-Friendly Streamlit Interface
Keywords:
Fake accounts, Social media, Machine learning, Random Forest, Streamlit application, Fraud detection, Ensemble methodsAbstract
The proliferation of social media platforms has led to the widespread issue of fake accounts, which are often used for malicious purposes such as spreading misinformation and conducting fraud. This paper presents a machine learning-based approach to detecting fake accounts across Twitter, Facebook, and Instagram. By employing advanced algorithms like Random Forest, AdaBoost, and XGBoost, the system achieves high accuracy in identifying fake accounts, with Random Forest reaching an accuracy of 97% on Facebook, 100% on Twitter, and 91.67% on Instagram. A user-friendly Streamlit application facilitates real-time predictions and user interaction. The results demonstrate the effectiveness of ensemble methods, particularly Random Forest, in accurately detecting fake accounts, providing a scalable and efficient solution to enhance social media integrity.
