Decision Support System for Vocational Training Suitability of Job Applicants Using Naïve Bayes

Nazwatul Husna (Universitas Royal)
Novica Irawati (Universitas Royal)
Sahren Sahren (Universitas Royal)

Jurnal IPTEK Bagi Masyarakat
Vol 6 No 2 (2026)

Selecting suitable vocational training for job applicants requires a consistent and efficient assessment process. This study develops a web-based decision support system using the Naïve Bayes algorithm to classify applicants’ training suitability. The model was evaluated using 233 historical records, consisting of 186 training data and 47 testing data, with a stratified hold-out scheme. System functionality was tested using Black Box Testing, while classification performance was assessed through a confusion matrix. The system correctly classified 44 of 47 test records, achieving 93.61% accuracy and 85% balanced accuracy. It identified all suitable applicants with 100% recall, while recall for unsuitable applicants was 70%. Because the target labels were derived from historical staff decisions rather than independently validated ground truth, the system should be used as a decision-support tool rather than a replacement for staff judgment.