Feature Selection for Identification of Fake Profiles on Facebook

Abstract

AbstractTechnology is advancing at a breakneck speed these days. Online Social Network (OSN), which has become a part of everyone’s life in terms of making new friends and keeping track of friends and their interests. Social networking sites make social life better, but there are many problems when using these Social Media sites, especially Facebook. Problems i.e., privacy, offline, hacking are mainly done through fake profiles. Researchers found that 20–40% of profiles on social networking sites such as Facebook are fake profiles. So, the problem is to build an accurate model to detect if a Facebook profile is a fake profile based on the user’s social activity using machine learning techniques. As it is an automatic detection technique, machine can make it easier for the sites to manage the huge number of profiles, which cannot be done manually. There are many previous works on the identification of fake profiles. So this paper proposes the minimal set of generic features to identify the fake profiles on Facebook and the study determines that minimized set of main features are significant in the detection of the fake accounts on Facebook.KeywordsOnline social networksFake profilesMachine learningClassificationFeatures based techniques