AI-Driven Talent Acquisition: Enhancing Recruitment and Hiring with Machine Learning and Large Language Models

Abstract

Artificial intelligence (AI) is transforming corporate recruitment by enabling data-driven decision-making and automating candidate evaluation. This study presents an AI-powered recruitment framework that integrates machine learning and large language models (LLMs) to enhance resume classification, candidate-job matching, and contextual hiring insights. The research evaluates multiple machine learning algorithms—Support Vector Machine (SVM), Logistic Regression, Random Forest, K-Nearest Neighbors (KNN), and an Ensemble model—across different feature extraction strategies. Textual features were first extracted from resumes using Term Frequency–Inverse Document Frequency (TF-IDF) and Word2Vec embeddings. Additional linguistic and structural features, including keyword presence, n-grams, and part-of-speech (POS) tagging, were then incorporated to form a combined feature set. Results show that the SVM achieved the highest performance using TF-IDF features alone (accuracy = 99.17%, F1 = 99.29%), while the Random Forest model outperformed all others with the combined feature set (accuracy = 99.6%, F1 = 99.64%). Feature importance analysis revealed that TF-IDF contributed the most to model performance, followed by n-grams, keyword presence, and POS features. The integration of GPT-4 further supported the system by providing contextual feedback, skills gap analysis, and training recommendations for recruiters. The findings demonstrate that combining linguistic, statistical, and semantic features enhances both predictive accuracy and interpretability, offering a scalable, AI-driven approach to modern corporate talent acquisition. Future work will extend the framework to incorporate real-time labor market data, thereby expanding its applicability to dynamic corporate environments.