Integrated Approach for Detecting Fake Profiles By Utilizing a Hybrid of K-Means Clustering with Some Supervised Machine Learning Algorithms

Our daily lives are significantly shaped by social media platforms such as Facebook, Twitter, Instagram, LinkedIn, and others, where people engage actively on a global scale. However, these platforms grapple with the challenge of fake profiles, which can be created by humans, bots, or automated systems. These deceptive accounts serve various malicious purposes, including spreading rumors, fake news, and engaging in fraudulent activities like identity theft and phishing. In this article, we aim to tackle this issue by utilizing the K-Means algorithm as an Unsupervised Machine Learning Algorithm to segment our dataset. Subsequently, we will employ Supervised Machine Learning Algorithms such as KNN, Logistic Regression, Bernoulli Naive Bayes, SVM, and Linear SVC to improve the accuracy of detecting fake profiles.