<![CDATA[Learning evaluation plays an important role in improving course quality, yet user reviews are generally unstructured and difficult to analyze manually. This study aims to analyze user sentiment toward Udemy course evaluations using the Support Vector Machine (SVM) algorithm. The dataset consisted of 50 balanced English-language reviews (26 positive and 24 negative) selected from approximately 500 scraped reviews, with sentiment labels determined based on the rating categories provided by the source website. The research process included text preprocessing, TF-IDF feature weighting, SVM classification using a linear kernel, and 10-fold cross-validation. Model performance was evaluated using accuracy, precision, recall, and F1-score. The proposed model achieved an accuracy of 82% and an average F1-score of 0.82, indicating that the combination of TF-IDF and SVM effectively classifies user sentiment. These findings can assist course providers in evaluating user feedback and improving learning quality. Keywords: Sentiment Analysis; Support Vector Machine; TF-IDF; Business Intelligence; Learning Evaluation Abstrak Evaluasi pembelajaran berperan penting dalam meningkatkan kualitas kursus, namun ulasan pengguna umumnya berupa data teks tidak terstruktur yang sulit dianalisis secara manual. Penelitian ini bertujuan menganalisis sentimen pengguna terhadap evaluasi pembelajaran pada platform Udemy menggunakan algoritma Support Vector Machine (SVM). Dataset terdiri atas 50 ulasan berbahasa Inggris yang seimbang (26 positif dan 24 negatif) dari sekitar 500 hasil scraping, dengan label sentimen ditentukan berdasarkan kategori rating pada situs sumber. Tahapan penelitian meliputi preprocessing teks, pembobotan TF-IDF, klasifikasi SVM berkernel linear, dan validasi menggunakan 10-fold cross validation. Kinerja model dievaluasi menggunakan accuracy, precision, recall, dan F1-score. Hasil penelitian menunjukkan accuracy sebesar 82% dan rata-rata F1-score sebesar 0,82, sehingga kombinasi TF-IDF dan SVM efektif mengklasifikasikan sentimen pengguna serta dapat mendukung peningkatan kualitas pembelajaran. Kata Kunci: Analisis Sentimen; Support Vector Machine; TF-IDF; Business Intelligence; Evaluasi Pembelajaran]]>
