ENHANCING ONLINE JOB VERIFICATION BY USING DEEP LEARNING

Abstract :

The proliferation of fraudulent job postings online poses a formidable obstacle for job seekers, demanding effective detection strategies. Current methodologies often falter in keeping pace with evolving deceitful tactics, resulting in compromised accuracy. This endeavor introduces an innovative Deep Learning paradigm, harnessing Dual Directional LSTM architecture to fortify counterfeit job identification. Through rigorous training and evaluation on a comprehensive dataset, our model surpasses contemporary benchmarks, achieving an impressive accuracy of 96.8%. Noteworthy precision (81.3%), recall (82.4%), and f1-score (81.1%) further underscore its efficacy. Moreover, our initiative endeavors to enhance accessibility by seamlessly integrating the model into a web server, facilitating real-time analysis of job postings for users.