Fake Social Media Profile Detection

Our project is based on the design and deployment of a high-end machine learning model for the identification of spurious social media profiles. The project uses large datasets with attributes of actual and spurious profiles, which are mixed up in a deep neural network model for efficient profile classification. Major contributions are the strategic attribute design of metrics like follower-to- following ratio and engagement metrics, as well as the application of SMOTE to address class imbalance. Model performance is rigorously tested using a range of metrics like accuracy, precision, recall, AUC, and F1-score. Also, the system design is optimized for real-world deployment with the modular backend implemented in FastAPI using Python and the dynamic and interactive frontend implemented using Node.js. This configuration not only provides high detection accuracy but also is simple to deploy and maintain, breaking new ground in the battle against online fake personas.