Fraud-Free Employment: Leveraging Gaussian Naive Bayes for Job Validation

This research study develops a reliable and efficient Naive Bayes-based framework for detecting fake job advertisements, utilizing optimized TF-IDF feature extraction methods. Materials and Methods: The proposed system uses Naive Bayes with TF-IDF to improve precision and recall while minimizing false detections, compared to the existing DNN framework, which employs multiple classifiers for accuracy and optimization. Results: The Naive Bayes model outperforms the Deep Neural Network model with an accuracy range of 87.9 % – 91.8 % (SD: 1.45429) versus 78.5 % – 82 % (SD: 1.51217), reduces processing latency to 1.70 seconds, and achieves statistical significance (p < 0.05). Conclusion: The Naive Bayes framework demonstrates superior accuracy, efficiency, and reliability, proving its effectiveness as a robust solution for identifying fraudulent job postings.