Article Per Year (5 Year)
p-Index From 2021 - 2026
1.573
P-Index
This Author published in this journals
All Journal JAIS (Journal of Applied Intelligent System) JURNAL PENGABDIAN KEPADA MASYARAKAT METHOMIKA: Jurnal Manajemen Informatika & Komputerisasi Akuntansi JOURNAL OF SCIENCE AND SOCIAL RESEARCH IT (INFORMATIC TECHNIQUE) JOURNAL JUDIMAS (Jurnal Inovasi Pengabdian Kepada Masyarakat) Jurnal IPTEK Bagi Masyarakat Jurnal Hasil Pengabdian Masyarakat (JURIBMAS) Jurnal Pengabdian Pada Masyarakat METHABDI Sabangka Abdimas Jurnal Pengabdian Masyarakat Sabangka
Linda Wahyuni
Universitas Potensi Utama
Religion Aerospace Engineering Agriculture, Biological Sciences & Forestry Arts Humanities Civil Engineering, Building, Construction & Architecture Computer Science & IT Decision Sciences, Operations Research & Management Economics, Econometrics & Finance Education Electrical & Electronics Engineering Engineering Environmental Science Immunology & microbiology Industrial & Manufacturing Engineering Languange, Linguistic, Communication & Media Mechanical Engineering Medicine & Pharmacology Physics Public Health Social Sciences Other

Published : 11 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 11 Documents
Search

 1   2 
KLASIFIKASI SENTIMEN BERBASIS MACHINE LEARNING TERHADAP PENONTON “WONDERLAND INDONESIA” PADA PLATFORM YOUTUBE Surya Darma; Linda Wahyuni; Nauval Alfarizi
JOURNAL OF SCIENCE AND SOCIAL RESEARCH Vol. 9 No. 2 (2026): April 2026
Publisher : Smart Education

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.54314/jssr.v9i2.6270

Abstract

Abstract: The purpose of this study is to analyze and classify audience sentiment toward the Wonderland Indonesia video on the YouTube platform using the Naive Bayes and Support Vector Machine (SVM) algorithms. The dataset consists of 10,000 comments collected through a scraping process. The data were then further processed using the Sastrawi library, including data cleaning, case folding, tokenization, stopword removal, and stemming. The TF-IDF method with a dimensional size of (9048, 5000) was used for feature representation. The data distribution indicates a class imbalance, where neutral (4,550) and positive (4,297) sentiments dominate compared to negative (201). The experimental results show that SVM outperforms Naive Bayes, achieving an accuracy of 93% compared to 86%. However, both models encounter difficulties in identifying the negative class. The use of imbalanced data handling techniques shows that SVM with class weighting improves the detection of minority classes without reducing overall accuracy. In contrast, Naive Bayes with SMOTE improves recall but reduces overall performance. The findings indicate that SVM with class weighting is the best-performing model for sentiment classification. Additionally, the results show that the majority of comments are positive, suggesting that audiences respond favorably to the Wonderland Indonesia content. Keywords: Sentiment Analysis, Naive Bayes, Support Vector Machine, YouTube, TF-IDF Abstrak: Tujuan penelitian ini adalah untuk menganalisis dan mengklasifikasikan sentimen penonton terhadap video Wonderland Indonesia yang ditampilkan di platform YouTube menggunakan algoritma Naive Bayes dan Support Vector Machine (SVM). Data yang digunakan terdiri dari 10.000 komentar yang dikumpulkan melalui proses scraping. Kemudian, data diproses lebih lanjut dengan menggunakan pustaka Sastrawi untuk membersihkan, menggabungkan case, tokenize, menghilangkan stopword, dan stemming. Metode TF-IDF dengan dimensi (9048, 5000) digunakan untuk mempresentasikan fitur. Distribusi data menunjukkan ketidakseimbangan kelas; sentimen netral (4.550) dan positif (4.297) mendominasi dibandingkan dengan negatif (201). Hasil pengujian menunjukkan bahwa SVM unggul dengan akurasi 93% dibandingkan Naive Bayes sebesar 86%. Namun, kedua model kesulitan menemukan kelas negatif. SVM dengan berat kelas dapat meningkatkan deteksi kelas minoritas tanpa mengurangi akurasi, seperti yang ditunjukkan oleh penggunaan teknik penanganan data tidak seimbang. Di sisi lain, Naive Bayes dengan SMOTE meningkatkan recall tetapi mengurangi performa keseluruhan. Hasil penelitian menunjukkan bahwa model terbaik untuk klasifikasi sentimen adalah SVM dengan berat kelas. Hasil penelitian juga menunjukkan bahwa sebagian besar komentar bersentimen positif, yang menunjukkan bahwa audiens menyukai konten Wonderland Indonesia. Kata Kunci: Analisis Sentimen, Naive Bayes, SVM, YouTube, TF-IDF
Co-Authors Abdul Meizar Afandi Maulana Ananda Fitriani Dewi Andika Wiratama Chairul Rizal Charles Bronson Harahap Chrairul Rizal Elida Tuti Siregar Elsa Aditya Erwin Ginting Fujiati Fujiati Hardianto Hardianto Iwan Fitrianto Rahmad Khairul Ummi Maftuhin Muhammad Eka Nauval Alfarizi Negoro, Wahyu Saptha Nidia Enjelita Saragih Rahmi Hayati Rika Rosnelly Rony, Zahara Tussoleha Soeheri Soeheri Supiyandi Supiyandi Surya Darma Wirhan Fahrozi Wirhan Fahrozi Yusuf Ramadhan Nasution