Online romance scams are widespread in the West, but there’s limited data-driven research on them. Scammers create fake profiles and interact with victims. Traditional spam filters are ineffective. We analyze scam profiles, their strategies, and victim traits. We’ve developed a system (97% accuracy) for early scam detection, aiming to protect victims and dating platforms. This advances existing research by incorporating idealized romantic beliefs into our machine-learning model. It offers automated solutions for dating site providers and users.

• DataDrivenResearch
• OnlineRomanceScams
• ScamDetection