Detecting Fraudulent Job Postings: A Machine Learning Approach

Abstract

The advent of fraudulent job advertising poses a serious danger to job searchers' confidence and safety in the age of online employment marketplaces. Our research project develops an advanced machine learning-based system to predict the legitimacy of job profiles, which tackles this urgent problem. Utilizing state-of-the-art Python libraries like scikit-learn, NLTK, and spaCy, together with knowledge from earlier research, our project uses a thorough preprocessing pipeline to extract useful features from textual input. By combining several machine learning methods, such as Random Forest, Multinomial Naive Bayes, and Logistic Regression, with a Voting Classifier framework, we develop an ensemble model that can reliably distinguish between real and fake job listings. Our study strives to give job searchers a trustworthy tool to navigate the internet job market safely and confidently by methodically assessing and improving our models.