<![CDATA[The education-job mismatch phenomenon remains a significant challenge for university graduates, where many individuals work in fields that are not aligned with their educational background and competencies. Career decision-making processes are also generally subjective and have not fully leveraged data-driven analysis. This study aims to design and implement a post-graduation career-fit prediction system based on resume screening using a machine learning approach. The proposed method employs two supervised classification algorithms, namely Random Forest and Support Vector Machine (SVM), with feature representation using TF-IDF based on n-grams on a dataset of 13,389 resumes. The results indicate that both models achieve strong performance; however, SVM outperforms Random Forest, achieving an accuracy of 84.39% and an F1-score of 83.36%, compared to 81.96% accuracy for Random Forest. Feature importance analysis reveals that technical skills, work experience, and field of study are the most influential factors in determining career fit. This study contributes a data-driven predictive approach to support more objective career decision-making for students and graduates.]]>
Copyrights © 2026
