Neural Network-Based Detection of Fraudulent Profiles In Social Media Platforms

Abstract

Social media platforms have become pervasive in
modern society, offering opportunities for
individuals to connect, share information, and
engage in various activities. However, the rise of
fraudulent activities, such as fake profiles, poses
significant challenges to the integrity and security of
these platforms. Traditional methods of detecting
fraudulent profiles often rely on manual inspection
or rule-based systems, which can be time-consuming
and ineffective in identifying sophisticated
fraudulent behavior. This study proposes a novel
approach using neural networks for the automated
detection of fraudulent profiles in social media
platforms. By leveraging the power of deep learning
techniques, the proposed system learns intricate
patterns and features from large-scale datasets,
enabling it to effectively distinguish between
genuine and fraudulent profiles. The neural network
model is trained on diverse sets of features,
including user behavior patterns, content
characteristics, network structure, and temporal
dynamics, to capture the complex nature of
fraudulent activities. Experimental results
demonstrate the efficacy of the proposed approach
in detecting fraudulent profiles with high accuracy
and efficiency. Compared to traditional methods,
the neural networkbased detection system achieves
superior performance in terms of precision, recall,
and F1-score. Moreover, the model exhibits
robustness against various evasion techniques
employed by fraudsters, making it suitable for realworld
deployment in social media platforms. Social
networking sites such as Facebook, Twitter,
histogram, etc. are extremely famous among people.
Users always interact with their friends via these
social networks sites or media. They share their
personal and public information using these social
networks. an immense number of people use social
networking sites due to their attractiveness. This
fame causes problems to the websites due to the
creation of fake accounts. The owners of fake
accounts pull out personal information about other
people and spread the fake data on social networks.
In our proposed plan, we propose machine learning
techniques such as Neural Networks and SVM for
detecting fake accounts on Facebook or Twitter, or
Twitter.