An Explainable and Fair AI-Based Recruitment Screening System Using Machine Learning with Fairness-aware Evaluation & Decision Optimization

Abstract

The recruitment procedure in organizations is a time consuming, subjective, and inefficient hiring system in the modern era due to the manual screening of large numbers of applications. This paper presents an AI-based Explainable recruitment framework. To overcome these challenges, this research develops a unified pipeline that integrates Natural Language Processing (NLP) and Machine Learning (ML). In this system, we use different techniques for effective candidate screening and automated results optimization. The intelligent system extracts specific information using TF-IDF and BERT embeddings related to hiring requirements from large numbers of resumes using NLP modules like text parsing analyzing resumes through targeted keywords. In this way the system converts unstructured data into useable structured features. By using different ML models, including XGBoost, logistic regression, random forest, SHAP + LIME, and fairness metrics for classification, accuracy, and applicant selection for the explanation and final hiring decision. These modules can substantially improve the accuracy and consistency of applicant selection. Accentuates the capability of AI-based systems to mutate traditional hiring practices into intelligent solutions.