Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{200742,
        author = {Yasmin R. Shaikh and Swati S. Hinge},
        title = {Real-Time Detection of Fraudulent Job Postings Using Machine Learning, Explainable AI and Flask Deployment},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {12},
        pages = {3808-3816},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=200742},
        abstract = {The proliferation of fraudulent job postings across online recruitment portals has created a mounting cybersecurity threat that harms job seekers financially and psychologically. Automated detection at scale is no longer optional. This paper presents a complete, production-ready machine learning framework that detects fraudulent job advertisements with 98.43% accuracy. Building on our previously published work [25], we introduce three substantive advances: (i) XGBoost combined with ADASYN oversampling as the new best-performing classifier, surpassing our earlier Random Forest baseline; (ii) feature importance analysis revealing "data entry" (score 0.0108) and "earn" (0.0090) as the dominant fraud-indicating linguistic patterns, advancing Explainable AI (XAI) for this domain; and (iii) a fully deployed Flask web application, JobGuard AI, providing real-time fraud predictions through an intuitive browser interface. The Employment Scam Aegean Dataset (EMSCAD) of 17,880 job postings is used throughout. TF-IDF with n-gram range (1,2) extracts 5,000 features. SMOTE and ADASYN independently correct the severe 20:1 class imbalance. All experimental results are reproduced and visualised in a Jupyter-based pipeline and validated through the deployed web system.},
        keywords = {fraudulent job detection, machine learning, NLP, TF-IDF, SMOTE, ADASYN, XGBoost, Random Forest, ensemble learning, EMSCAD dataset, explainable AI, Flask deployment, online recruitment security.},
        month = {May},
        }