BeNet: Hybrid Deep Learning-Driven Behavioral Analysis for Career Decision-Making

BeNet is a hybrid deep learning–driven behavioral analysis system designed to support career decision-making by predicting suitable professional roles based on academic performance, technical competencies, and behavioral attributes. The system integrates traditional machine learning and deep learning models within a secure web-based framework developed using Flask and MySQL. BeNet enables users to register, authenticate, upload structured datasets, evaluate multiple predictive models, and obtain personalized career recommendations through an interactive interface. The framework employs a diverse set of classification models, including CatBoost, Multi-Layer Perceptron (MLP), 1D Convolutional Neural Network (1D-CNN), a Hybrid Random Forest with 1D-CNN architecture, and a stacking-based ensemble classifier. These models analyze key indicators such as higher secondary education scores, aptitude, problem-solving ability, programming skills, design proficiency, data-structure coding knowledge, participation in technical events, extracurricular activities, and institutional engagement. Performance evaluation is conducted using accuracy, precision, recall, and macro-averaged F1-score to ensure reliable and balanced predictions across career categories. BeNet demonstrates strong predictive performance, with ensemble and gradient-boosting models achieving high classification accuracy. The system outputs interpretable career recommendations, including roles such as Software Developer, Data Analyst, Machine Learning Engineer, UI/UX Designer, and Technical Support. By combining behavioral analytics with hybrid learning architectures, BeNet provides an effective decision-support tool for data-driven career guidance.

Keywords: Career Decision Support, Behavioral Analysis, Hybrid Deep Learning, CatBoost, 1D Convolutional Neural Network, Ensemble Learning, Flask Application