Machine Learning Based Online Scam Reportage System for Security Agencies

The use of technology in our daily lives has led to an increase in cybercrime worldwide. Recently, cybercrime cases have been on the rise. This project aimed to develop a mobile application for reporting online scams aiding machine learning. The mobile application allows users to simultaneously report cybercrime incidents to security agency, attach evidence to the report. The application was developed using flutter, dart, scikit-learn, python and the development process followed the agile development methodology. The application was tested for functionality, usability, and security, and it was found to be fully functional, highly usable, and secure. Machine learning was integrated with the mobile app to analyze collected reports and identify patterns in scammer tactics, communication styles, and potential red flags. Random forest classifier was used to train the model. The Random Forest Classifier performed well in detecting scam communications, achieving an accuracy of 91.12% during testing, key metrics used to evaluate the model include precision, accuracy, recall, and F1 score, all of which demonstrated the system’s effectiveness in identifying online scams. The mobile application for reporting online scams has the potential to improve the reporting and investigation of cybercrime in Nigeria