NLP-Powered Resume Matching For Recruitment

More recently, recruitment has largely relied on the automation of the matching process between candidates and job roles. In this regard, this paper focuses on the development of a resume parser application utilizing NLP techniques, PDF text extraction, and machine learning-based evaluation of resumes according to job descriptions. In developing the application using the Flask framework, users are allowed to upload resume and job description files in PDF format. The system automatically extracts the text, preprocesses it, and performs the task of skill matching. It also computes a semantic similarity score based on term frequency-inverse document frequency and cosine similarity techniques. Using a trained machine learning model, the application predicts a binary job fit score based on its semantic similarity and skills matching metrics scores. This paper outlines the design, implementation, and evaluation of the system, and it indeed has the potential to assist the recruiters in pre-screening the candidates.