<![CDATA[ABSTRAK Pemilihan kompetensi keahlian yang tidak tepat di SMK seringkali memicu kendala akademik dan ketidaksesuaian minat bagi siswa baru. Penelitian ini bertujuan untuk menguji tingkat akurasi model Certainty Factor (CF) pada sistem pakar rekomendasi jurusan SMK guna mengatasi ketidakpastian dalam penilaian minat dan bakat. Mengingat keterbatasan akses data lapangan, penelitian ini menerapkan pendekatan studi literatur dengan basis pengetahuan sistem yang diakuisisi dari regulasi kurikulum resmi Kemendikbudristek dan pedoman bimbingan konseling. Evaluasi performa klasifikasi model diuji secara ketat menggunakan metode Confusion Matrix terhadap 30 skenario data uji sintetis yang dirancang secara sistematis dari berbagai variasi nilai akademik serta indikator minat. Hasil pengujian menunjukkan bahwa model Certainty Factor mampu mengimplementasikan logika penjurusan dari literatur dengan sangat baik, menghasilkan tingkat Akurasi 93.3%, rata-rata presisi 93.3%, dan rata-rata recall 93.3%. Kesimpulan dari penelitian ini membuktikan bahwa pengujian menggunakan Confusion Matrix berhasil memvalidasi konsistensi kode program dan rumus matematika CF dalam meniru aturan penjurusan normatif secara presisi. Sistem pakar ini dinyatakan valid dan layak dikembangkan lebih lanjut sebagai instrumen bantu (decision support tool) bagi guru bimbingan konseling maupun calon siswa baru, meskipun proses evaluasinya berbasis pada skenario sintetis. ABSTRACT Inappropriate selection of vocational competencies in vocational high schools (SMK) often triggers academic difficulties and interest mismatches among new students. This study aims to evaluate the accuracy level of the Certainty Factor (CF) model within an expert system for recommending SMK majors to address uncertainty in interest and talent assessment. Due to limited access to field data, this research employs a literature review approach, where the system's knowledge base is acquired from the official curriculum regulations of the Ministry of Education, Culture, Research, and Technology (Kemendikbudristek) and guidance counseling manuals. The classification performance of the model is rigorously evaluated using the Confusion Matrix method against 30 synthetic test data scenarios systematically designed from various academic grades and interest indicators. The evaluation results indicate that the Certainty Factor model successfully implements the majoring logic derived from the literature, yielding an accuracy rate of 93.3%, a macro average precision of 93.3%, and a macro average recall of 93.3%. In conclusion, this study demonstrates that the Confusion Matrix evaluation successfully validates the consistency of the program code and the CF mathematical formula in precisely replicating normative majoring rules. This expert system is deemed valid and viable for further development as a decision support tool for guidance counseling teachers and prospective students, despite the evaluation process being reliant on synthetic scenarios.]]>
Copyrights © 2026
