International Journal of Scientific Research and Engineering Development

In today's rapidly evolving job market, connecting the right candidates to the right opportunities remains a persistent challenge. Traditional job portals rely on keyword-based search and manual filtering, failing to understand the semantic relationships between skills and job requirements. This paper presents a Skill-Based Job Recommendation and Matching System that leverages Natural Language Processing (NLP) and Machine Learning to intelligently extract skills from resumes and job descriptions, compute semantic similarity, and deliver highly personalised job recommendations. The system uses Named Entity Recognition (NER) with spaCy to extract structured skills from free-form text, encodes them using TF-IDF and BERT (Sentence Transformers), and computes cosine similarity for precise job-candidate matching. A hybrid recommendation engine combining Content-Based Filtering (CBF) and Collaborative Filtering (CF) ranks top-N jobs per candidate. The platform further provides skill gap analysis with course recommendations and a recruiter-facing candidate ranking portal. Evaluation results demonstrate a Precision@5 of 88%, Recall@10 of 85%, NDCG of 0.91, and Match Accuracy of 93%.

K.Sandha Kavishman, Mr.R.Ramakrishnan,"Skill-Based Job Recommendation and Matching System" International Journal of Scientific Research and Engineering Development, V9(3): Page(1418-1424) May-June 2026. ISSN: 2581-7175. www.ijsred.com. Published by Scientific and Academic Research Publishing.