Pengaruh Integrasi Explainable MCDM Terhadap Kualitas Sistem Pendukung Keputusan Seleksi Peserta Magang PT Semen Indonesia

The Impact of Explainable MCDM Integration on the Quality of the Decision Support System for Internship Participant Selection at PT Semen Indonesia

Seleksi kandidat magang di perusahaan sering menghadapi tantangan seperti penilaian manual, proses yang memakan waktu, dan kurangnya transparansi dalam pengambilan keputusan pemeringkatan, yang menyebabkan subjektivitas dan penurunan kualitas keputusan. Studi ini mengembangkan Sistem Pendukung Keputusan (SPK) berdasarkan penjelasan Multi-Criteria Decision Making (MCDM) yang mengintegrasikan Analytical Hierarchy Process (AHP) dan Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) dengan Explainable Artificial Intelligence (XAI), yang diimplementasikan menggunakan Laravel, Livewire, PostgreSQL, dan API Groq. Sistem ini mengevaluasi kandidat berdasarkan lima kriteria: IPK, Jumlah Sertifikasi, Skor Evaluasi CV, Kesesuaian Persyaratan Pendidikan, dan Skor Kesesuaian Bidang. Hasil evaluasi menunjukkan bahwa sistem tersebut mencapai skor kualitas ISO 9126 sebesar 1,58 (sesuai), korelasi Spearman sebesar 0,9643 yang menegaskan kesesuaian yang kuat dengan peringkat SDM manual, konsistensi peringkat 100%, skor penjelasan dan transparansi masing-masing sebesar 92,50% dan 91,25%, waktu pemrosesan rata-rata 16,40 detik, dan Skor Lapisan XAI sebesar 92,04% (sangat baik). Hasil ini menegaskan bahwa sistem tersebut secara efektif mendukung seleksi kandidat magang dengan cara yang lebih objektif, akurat, transparan, dan efisien.

Internship candidate selection in companies often faces challenges such as manual assessment, time-consuming processes, and lack of transparency in ranking decisions, leading to subjectivity and reduced decision quality. This study develops a Decision Support System (DSS) based on Explainable Multi-Criteria Decision Making (MCDM) integrating Analytical Hierarchy Process (AHP) and Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) with Explainable Artificial Intelligence (XAI), implemented using Laravel, Livewire, PostgreSQL, and Groq API. The system evaluates candidates across five criteria: GPA, Certification Count, CV Evaluation Score, Education Requirement Match, and Field Match Score. Evaluation results show that the system achieved an ISO 9126 quality score of 1.58 (appropriate), a Spearman correlation of 0.9643 confirming strong agreement with manual HR rankings, 100% ranking consistency, explainability and transparency scores of 92.50% and 91.25% respectively, an average processing time of 16.40 seconds, and an XAI Layering Score of 92.04%(very good). These results confirm that the system effectively supports internship candidate selection in a more objective, accurate, transparent, and efficient.