Analisis Klasterisasi Profil Keamanan Siber Mahasiswa untuk Penentuan Strategi Edukasi

Abstract

Mahasiswa perguruan tinggi merupakan kelompok rentan terhadap ancaman keamanan siber akibat tingginya intensitas penggunaan teknologi digital. Penelitian ini bertujuan mengukur tingkat literasi keamanan siber, mengelompokkan profil mahasiswa, mengidentifikasi karakteristik setiap klaster, dan menyusun rekomendasi edukasi adaptif. Survei dilakukan terhadap 204 mahasiswa IPB University yang belum mengambil mata kuliah keamanan informasi, terdiri dari 130 mahasiswa IT dan 74 mahasiswa Non-IT, menggunakan instrumen berbasis skala Likert empat poin pada dimensi Knowledge dan Attitude. Uji reliabilitas dengan Cronbach's Alpha menghasilkan nilai 0,912 (Knowledge IT), 0,797 (Attitude IT), 0,768 (Knowledge Non-IT), dan 0,770 (Attitude Non-IT), seluruhnya memenuhi ambang batas reliabel. Algoritma K-Means Clustering dengan Silhouette Score (K=2 optimal: IT=0,442; Non-IT=0,414) menghasilkan dua klaster pada masing-masing kelompok. Klaster IT Literasi Tinggi (n=68) memiliki skor kumulatif 3,182, sedangkan Klaster IT Literasi Medium (n=62) memperoleh 2,254. Pada Non-IT, Klaster Literasi Tinggi (n=37) memperoleh 3,576 dan Klaster Literasi Medium (n=37) memperoleh 2,966. Analisis tujuh tema literasi menunjukkan Social Engineering sebagai kelemahan utama kelompok IT, sementara WiFi & Safe Browsing menjadi kelemahan utama klaster Non-IT Literasi Medium. Rekomendasi edukasi disusun secara terdiferensiasi per klaster untuk meningkatkan efektivitas intervensi keamanan siber di lingkungan perguruan tinggi

University students are a vulnerable group to cybersecurity threats due to their high intensity of digital technology use. This study aims to measure students' cybersecurity literacy levels, cluster their profiles, identify cluster characteristics, and develop adaptive educational recommendations. A survey was conducted on 204 IPB University students who had not taken information security courses, comprising 130 IT students and 74 Non-IT students, using a four-point Likert scale instrument measuring Knowledge and Attitude dimensions. Reliability testing using Cronbach's Alpha yielded values of 0.912 (IT Knowledge), 0.797 (IT Attitude), 0.768 (Non-IT Knowledge), and 0.770 (Non-IT Attitude), all meeting the reliability threshold. K-Means Clustering with Silhouette Score (optimal K=2: IT=0.442; Non-IT=0.414) produced two clusters for each group. The IT High Literacy cluster (n=68) achieved a cumulative score of 3.182, while the IT Medium Literacy cluster (n=62) scored 2.254. For Non-IT students, the High Literacy cluster (n=37) scored 3.576 and the Medium Literacy cluster (n=37) scored 2.966. Analysis of seven literacy themes revealed Social Engineering as the primary weakness for IT groups, while WiFi and Safe Browsing was the main weakness for Non-IT Medium Literacy cluster. Differentiated educational recommendations were developed per cluster to improve the effectiveness of cybersecurity interventions in higher education settings.