EXPLAINABLE AND DYNAMIC KNOWLEDGE GRAPH-AUGMENTED RETRIEVAL GENERATION FOR FAIR AND EFFICIENT APPLICANT-JOB MATCHING

The increasing reliance on AI-driven systems for recruitment poses challenges in fairness, transparency, and adaptability. This paper introduces a novel framework that integrates dynamic knowledge graphs (DKGs) with retrieval-augmented generation (RAG) to improve the accuracy, explainability, and fairness of applicant-job matching. By continuously updating knowledge graphs with real-time information and leveraging RAG for context-aware recommendations, the system addresses the limitations of traditional recommender systems in HR. Extensive experiments using real-world datasets demonstrate superior performance in terms of precision, recall, fairness metrics, and explanation quality compared to baseline models. The framework also provides interpretable, human-readable justifications for recommendations, enhancing recruiter and applicant trust.