An AI-assisted scam detection and notification system
Shabir Ali, Shankranand Sarswati, Aashish Ranjan · 2025
User security and safety are compromised on email, WhatsApp, and SMS due to this increase in spam and scam messages. To curtail the challenge, this project employs the successful text classification approach known as the Naive Bayes algorithm to formulate a spam filter system. Using a dataset consisting of 5,300 correctly classified messages, the algorithm simply classifies text as either “spam” or “ham”-genuine once it has seen the training process. Currently, it uses a trained model to process data from a CSV file and correctly identify spam. The initiative hopes to integrate pertinent APIs to enable real-time scam detection across several platforms, including Gmail, Snapchat, Instagram, and WhatsApp. This extension will notify users of possible frauds and enable automatic text analysis. A web and mobile interface will also be developed for permission control and displaying the findings to the users. Cloud architecture, either from AWS or Google Cloud, will be used for scalability, and the focus will be on data security and user privacy. To enhance protection against spam and scams on various communication channels, future developments will include larger datasets, more advanced algorithms, and seamless cross-platform interaction.