Job Tier Classification in Applicant Tracking Systems Utilizing One-Dimensional CNN

The increasing volume of job applications has made resume screening a critical yet challenging task for recruiters.Traditional Applicant Tracking Systems (ATS) often rely on keyword-based filtering, which may lead to inefficiencies and biases in candidate evaluation.This study proposes a One-Dimensional Convolutional Neural Network (1D CNN) model for job tier classification, utilizing TF-IDF embeddings and Named Entity Recognition (NER) to extract essential features from resumes.We specifically sought to evaluate the model's performance against conventional recruitment methods by comparing classification accuracy on labeled datasets comprising 1,444 resumes.The findings indicate that our approach achieved a 60.98% improvement in classification accuracy over traditional methods, with education being identified as the most significant predictor of job tier classification.However, limitations related to potential dataset biases and the reliance on automated evaluations are acknowledged.This research illustrates the potential for deep learning techniques to enhance recruitment automation, offering a scalable solution that may help reduce bias in resume classification within ATS.