Job Description and Resume Matching System Using NLP

The paper proposes a deep learning-based framework aimed at automating the process of matching job descriptions with resumes. The authors employ a BERT (Bidirectional Encoder Representations from Transformers) model, which is fine-tuned for semantic relevance assessment, in scoring multiple resumes for a given job description. It also embeds all-embracing text preprocessing methods as well as methods for feature extraction and classification to increase the quality of matching. As such, the new method reduces the time taken to evaluate multiple CVs thereby improving hiring processes and making better hiring decisions. Evaluation of the proposed models performed on actual data yielded good results, proving the idea of recruiting assistants based on NLP (Natural Language Processing) models viable. This work demonstrates the effectiveness of applying BERT for semantic similarity tasks of a human resource management system. Our combination of Job Description and Resume makes the closest match with a similarity of 97%.