drveronika/x_fake_profile_detection

Tasks

• image-classification

Citation text from the card

@Article{electronics13132542,
	AUTHOR = {Dracewicz, Weronika and Sepczuk, Mariusz},
	TITLE = {Detecting Fake Accounts on Social Media Portals—The X Portal Case Study},
	JOURNAL = {Electronics},
	VOLUME = {13},
	YEAR = {2024},
	NUMBER = {13},
	ARTICLE-NUMBER = {2542},
	URL = {https://www.mdpi.com/2079-9292/13/13/2542},
	ISSN = {2079-9292},
	ABSTRACT = {Today, social media are an integral part of everyone’s life. In addition to their traditional uses of creating and maintaining relationships, they are also used to exchange views and all kinds of content. With the development of these media, they have become the target of various attacks. In particular, the existence of fake accounts on social networks can lead to many types of abuse, such as phishing or disinformation, which is a big challenge nowadays. In this work, we present a solution for detecting fake accounts on the X portal (formerly Twitter). The main goal behind the developed solution was to use images of X portal accounts and perform image classification using machine learning. As a result, it was possible to detect real and fake accounts and indicate the type of a particular account. The created solution was trained and tested on an adequately prepared dataset containing 15,000 generated accounts and real X portal accounts. The CNN model performing with accuracy above 92% and manual test results allow us to conclude that the proposed solution can be used to detect false accounts on the X portal.},
	DOI = {10.3390/el

All datasets marked Permissive or share-alike | JSON