Dual-Phase Resume Parsing and Attrition Prediction System

This paper introduces a high-parameterized resume parsing model that resolves two vital aspects of human resource analytics: prediction of employee attrition and CV shortlisting by automated means. The initial step applies conventional supervised learning models such as Logistic Regression, Random Forest, and Gradient Boosting for predicting employee attrition. The second phase deals with resume parsing through Natural Language Processing (NLP) by applying TF-IDF vectorization and Cosine Similarity for matching job descriptions with resumes. The attrition prediction dataset is real and public, while synthetic job descriptions and resumes are used for CV shortlisting based on privacy restrictions. Evaluation criteria include confusion matrix, cross-validation, and grid search optimization.