<![CDATA[Abstrak - Perkembangan teknologi informasi mempermudah akses layanan publik, termasuk aplikasi Parak Acil Online yang dikembangkan oleh Pemerintah Kota Banjarmasin untuk pengurusan dokumen administrasi. Sejak diluncurkan, aplikasi ini telah digunakan oleh puluhan ribu warga. Penelitian ini bertujuan untuk menganalisis sentimen pengguna terhadap aplikasi dan mengevaluasi performa Support Vector Machine dalam klasifikasi ulasan. Metode penelitian yang digunakan adalah Support Vector Machine untuk mengklasifikasikan ulasan pengguna. Hasil analisis menunjukkan bahwa algoritma Support Vector Machine (SVM) dalam mengklasifikasikan data ulasan mendapatkan akurasi tertinggi pada pembagian data latih dan data uji 70:30 sebesar 85,1%, presisi 78,2%, dan recall 97,2%. Dari klasifikasi dan visualisasi, didapatkan kata-kata yang sering muncul pada sentimen positif yaitu “good”, “easy”, dan “helpful” serta kata-kata yang sering muncul pada sentimen negatif yaitu “difficult”, “take” dan “feature”. Sentimen masyarakat terhadap aplikasi Parak Acil Online menunjukkan bahwa mayoritas ulasan masyarakat terhadap aplikasi ini bersifat positif, dan performa analisis sentimen menggunakan metode Support Vector Machine yang digunakan dalam penelitian ini terbukti efektif dalam mengklasifikasikan sentimen dari ulasan pengguna. Diharapkan penelitian ini dapat membantu pengembang dan pemangku kebijakan dalam meningkatkan kualitas aplikasi Parak Acil Online serta memahami kebutuhan masyarakat.Kata kunci: Analisis Sentimen, Aplikasi Parak Acil Online, Support Vector Machine, Textblob. Abstract - The advancement of information technology has facilitated access to public services, including the Parak Acil Online application developed by the Banjarmasin City Government for managing administrative documents. Since its launch, this application has been used by tens of thousands of residents. This study aims to analyze user sentiment towards the application and evaluate the performance of Support Vector Machine (SVM) in classifying reviews. The research method used is Support Vector Machine (SVM) to classify user reviews. The analysis results show that the Support Vector Machine (SVM) algorithm achieves the highest accuracy in classifying review data with a 70:30 train-test split, reaching 85.1% accuracy, 78.2% precision, and 97.2% recall. Classification and visualization reveal that frequently occurring words in positive sentiment include "good," "easy," "helpful," and "fast," while frequently occurring words in negative sentiment include "difficult," "document," "take," and "feature." The sentiment of the public towards the Parak Acil Online application indicates that the majority of reviews are positive. The performance of sentiment analysis using the Support Vector Machine method employed in this study has proven effective in classifying sentiment from user reviews. It is hoped that this research can assist developers and policymakers in improving the quality of the Parak Acil Online application and understanding community needs.Keywords: parak acil online application, sentiment analysis, support vector machine, textblob.]]>
Copyrights © 2024
