thesis-resume-to-job-description-matching
PublicTraditional keyword-based job matching overlooks context. My MSc Thesis proposes a 3-stage NLP pipeline: - 1. HDBSCAN clustering of JDs by semantics, 2. cosine similarity for document-level matching, 3. skillset similarity via phrase extraction. The system ranks top-10 jobs using averaged contextual scores, offering more precise recommendations