<![CDATA[Privasi data merupakan aspek penting dalam aktivitas digital, terutama bagi mahasiswa yang aktif menggunakan berbagai platform daring. Penelitian ini bertujuan menganalisis tingkat kesadaran privasi data mahasiswa menggunakan algoritma Naïve Bayes. Data primer dikumpulkan melalui kuesioner Google Form yang berisi 13 indikator kesadaran privasi dan disebarkan melalui media sosial dengan teknik voluntary response sampling. Sebanyak 56 mahasiswa berpartisipasi sebagai sampel penelitian. Pengolahan data mengikuti tahapan Knowledge Discovery in Database (KDD), meliputi seleksi data, pembersihan, transformasi, pemodelan, serta evaluasi. Transformasi dilakukan dengan menghitung skor total per responden dan mengelompokkan tingkat kesadaran ke dalam kategori “Tinggi” dan “Standar” menggunakan cut-off empiris untuk menjaga keseimbangan kelas. Analisis klasifikasi dilakukan menggunakan algoritma Naïve Bayes melalui aplikasi Orange Data Mining, dengan evaluasi menggunakan Test and Score serta Confusion Matrix. Hasil penelitian menunjukkan bahwa model mampu mengklasifikasikan tingkat kesadaran privasi dengan akurasi 91.1%, precision 92.6%, recall 91.1%, F1-score 91.5%, AUC 0.976, dan MCC 0.738. Temuan ini menunjukkan bahwa Naïve Bayes efektif dalam mengenali pola kesadaran privasi mahasiswa dan layak digunakan sebagai dasar pengembangan program edukasi privasi data di lingkungan perguruan tinggi. Data privacy is a critical aspect of digital activity, particularly for university students who frequently engage with online platforms. This study aims to analyze students’ awareness of data privacy using the Naïve Bayes classification algorithm. Primary data were collected through a Google Form questionnaire consisting of 13 indicators of privacy awareness and distributed via social media using a voluntary response sampling technique. A total of 56 students participated in this study. Data processing followed the Knowledge Discovery in Database (KDD) stages, including data selection, cleaning, transformation, modeling, and evaluation. The transformation process involved calculating the total awareness score for each respondent and categorizing awareness levels into “High” and “Standard” using an empirical cut-off to maintain class balance. The Naïve Bayes algorithm was applied using the Orange Data Mining application, with performance evaluated through the Test and Score and Confusion Matrix tools. The results indicate that the model performed effectively, achieving an accuracy of 91.1%, precision of 92.6%, recall of 91.1%, F1-score of 91.5%, AUC of 0.976, and MCC of 0.738. These findings demonstrate that Naïve Bayes is suitable for analyzing student privacy awareness patterns and can serve as a foundation for designing educational interventions to improve privacy literacy in academic environments.]]>
