Analisis sentimen youtube terhadap komentar baik pada music video "Polyphia - playing God” dengan metode support vector machine

Lim, Nicholas (2024) Analisis sentimen youtube terhadap komentar baik pada music video "Polyphia - playing God” dengan metode support vector machine. Bachelor thesis, Universitas Pelita Harapan.

Preview	Text (Title) Title.pdf Available under License Creative Commons Attribution Non-commercial Share Alike. Download (40kB) | Preview
Preview	Text (Abstract) Abstract.pdf Available under License Creative Commons Attribution Non-commercial Share Alike. Download (200kB) | Preview
Preview	Text (ToC) ToC.pdf Available under License Creative Commons Attribution Non-commercial Share Alike. Download (142kB) | Preview
Preview	Text (Chapter1) Chapter1.pdf Available under License Creative Commons Attribution Non-commercial Share Alike. Download (261kB) | Preview
	Text (Chapter2) Chapter2.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (396kB)
	Text (Chapter3) Chapter3.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (468kB)
	Text (Chapter4) Chapter4.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (631kB)
	Text (Chapter5) Chapter5.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (168kB)
Preview	Text (Bibliography) Bibliography.pdf Available under License Creative Commons Attribution Non-commercial Share Alike. Download (192kB) | Preview
	Text (Appendices) Appendices.pdf Restricted to Repository staff only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (929kB)

Abstract

Penggunaan media sosial pada masyarakat, dengan maksud yang berbeda pada setiap orang teruslah meningkat. Pada Juni 2022, lebih dari 500 jam video telah diunggah tiap menitnya dan terdapat 4,95 milliar orang aktif bulanan di seluruh dunia. Setiap harinya pengguna Youtube menggunakan Youtube dari seluruh dunia. Data komentar dari Youtube akan digunakan untuk menentukan tingkat sentimen positif dan negatif pada Youtube dengan menggunakan metode Support Vector Machine untuk mengevaluasi performa akurasi keseluruhan, precision, recall dan f1-score. Dalam penelitian ini, terdapat tiga faktor yang dikelompokkan, yaitu faktor negatif, faktor positif dan faktor netral. klasifikasi dilakukan melalui beberapa tahap, yaitu pre-processing data, data training, data testing, dan evaluasi. Setelah melakukan evaluasi pada algoritma Support Vector Machine, diperoleh hasil tertinggi dengan akurasi keseluruhan mencapai 97%, precision negatif 99%, precision netral 95%, precision positif 98%, recall negatif 90%, recall netral 99%, recall positif 98%, dan f1-score negatif 94%, f1-score netral 97%, f1-score positif 98%. Hasil evaluasi yang tinggi menunjukkan bahwa model SVM yang digunakan memiliki kinerja yang baik dalam mengklasifikasikan sentimen komentar Youtube ke dalam tiga kategori yang ditentukan (negatif, netral, positif). Akurasi keseluruhan yang tinggi (97%) bersama dengan precision, recall, dan F1-score yang tinggi untuk masing-masing kategori sentimen menunjukkan bahwa model ini efektif dalam mengidentifikasi dan membedakan antara komentar negatif, netral, dan positif. / The use of social media in society, with different purposes for each person, continues to increase. As of June 2022, more than 500 hours of video were uploaded every minute and there were 4.95 billion monthly active people worldwide. Every day Youtube users use Youtube from all over the world. Comment data from YouTube will be used to determine the level of positive and negative sentiment on YouTube using the Support Vector Machine method to produce overall accuracy, precision, recall and f1-score performance. In this research, there are three factors that are believed to be, namely negative factors, positive factors and neutral factors. classification is carried out through several stages, namely data pre-processing, data training, data testing, and evaluation. After evaluating the Support Vector Machine algorithm, the highest results were obtained with overall accuracy reaching 97%, negative precision 99%, neutral precision 95%, positive precision 98%, negative recall 90%, neutral recall 99%, positive recall 98%, and negative f1-score 94%, neutral f1-score 97%, positive f1- score 98%.The high evaluation results show that the SVM model used has good performance in classifying YouTube comment sentiment into three specified categories (negative, neutral, positive). High overall accuracy (97%) along with high precision, recall, and F1-score for each sentiment category indicate that the model is effective in identifying and differentiating between negative, neutral, and positive comments.

Actions (login required)

View Item