The online social networking phenomenon has grown tremendously over the last twenty years. As social networking platforms have evolved, numerous online activities have emerged that have captured the attention of a large number of users.People increasingly rely on the credibility of the information presented on Online Social Networks (OSN). Conversely, online social networks have experienced a rise in the number of compromised, false accounts, scam profiles that do not correspond to real individuals.OSN operators are now using a variety of resources to detect such kind of scam profiles and accounts.Scammers in OSN are taking advantage of this for performing various OSN frauds. It is difficult to detect scammers due to the wide range of OSN platforms and the variety of OSN frauds. In this paper, an effort has been made to detect a scammer by designing a scammer detection model which will blacklist scammer profiles through user profile-based features. The proposed approach also differentiates between the scammer and real profiles. The experimental result and analysis show that the proposed model demonstrates better performance compared to other competing models, achieving an accuracy and f1 score of 98.75% and 97.95%, respectively for the dataset created for the study. This work aims to increase early-stage detection of scammers in dating frauds, compromised accounts, and fake profiles to provide safety to women and society.

Scammer profiles, Online social network,OSN frauds, Scammer detection model, social threats, Compromised accounts, Fake profiles, Dating Fraud, Machine Learning.