Re-Fake: Fake Account Classification in OSN Using RNN

Abstract

Online Social Network (OSN) is an application for enabling public communication and sharing information. However, the fake account in the OSN can spread false information with an unknown source. It is a challenging task to detect malicious accounts in a large OSN system. The existence of fake accounts or unknown accounts on OSN can be a severe issue in data privacy-preserving. Various communities have proposed many techniques to deal with fake accounts in OSN, including rule-based black-white technique until learning approaches. Therefore, in this study we propose a classification model using the RNN to detect fake accounts accurately and effectively. We conduct this study in several steps, including gathering datasets, pre-processing, extraction, training our models using RNN. Based on the experiment result, our proposed model can produce a higher accuracy than the conventional learning model