Real And Fake Job Post Detection Using Machine Learning

Abstract

The transition to online platforms has streamlined processes, reducing manual efforts. Job postings, now predominantly online, offer companies a broader reach for talent acquisition. However, amidst legitimate postings, fraudulent ones exist. This study aims to distinguish between real and fake job postings using machine learning techniques with optimal accuracy. Employing various data mining methods and classification algorithms—Logistic Regression, KNN, Decision Tree, XGBoost, Support Vector, Random Forest, and Multilayer Perceptron—we predict job authenticity. Supervised machine learning techniques guide dataset analysis, including variable identification and univariate, bivariate, and multivariate analyses, alongside handling missing values. Comprehensive data validation, cleaning, preparation, and visualization are conducted. Our experimentation, utilizing the Employment Scam Aegean Dataset (EMSCAD) with 17,881 samples, informs our approach. Ultimately, this research aims to enhance job seekers' security by effectively identifying fraudulent job postings. remove the writing issues. Key Words: Logistic Regression, KNN, Decision Tree, XGBoost, Support Vector, Random Forest, and Multilayer Perceptron.