Analisis Sentimen mengenai Vaksin Booster pada Komentar Video Youtube dengan Menerapkan Metode Support Vector Machine
Total View This Week0
Total View This Week0
Institusion
Institut Teknologi Kalimantan
Author
Tri Irawan, Hangga
Subject
QA75 Electronic computers. Computer science
Datestamp
2022-07-19 08:42:21
Abstract :
Social media provides easy access to social networking. One of them is Youtube. Youtube can provide a variety of information, entertainment, and news that can be taken into consideration in determining public opinion. The topic that is currently being discussed is the vaccine booster policy in Indonesia. To find out public opinion about this, analyzing sentiment on Youtube video commentary data regarding the vaccine booster policy is necessary. In analyzing sentiment, a suitable classification method is needed so that the polarity of the comment data is in the true polarity class. The raw data resulting from data retrieval from Youtube comments needs to be cleaned, intending to reduce noise in the data, such as deleting junk comments. So in this study, additional cleaning was carried out which was more adapted to the Youtube comment pattern. The classification method used is the Support Vector Machine (SVM) method. SVM is a classification method that has good performance in mapping data according to class and can work well on high-dimensional data. In the SVM method, it is necessary to select a kernel to train the model. In this study, experiments were carried out on both kernels, namely the linear kernel and the RBF. As for the results of the study, the model with the best performance was obtained based on testing with 3929 training data and 364 test data. Accuracy, precision, recall, and f1-score were obtained respectively 0.86, 0.85, 0.82, and 0.83. This model is built with a linear kernel, while the RBF kernel has lower evaluation results regarding accuracy performance and computation time. Then the results of the sentiment analysis regarding the booster vaccine were obtained, namely, from 364 comments, it was predicted that 210 negative comments and 154 positive comments were obtained. It can be concluded that public sentiment, especially Youtube users, tends to reject the vaccine booster policy.
Social media provides easy access to social networking. One of them is Youtube. Youtube can provide a variety of information, entertainment, and news that can be taken into consideration in determining public opinion. The topic that is currently being discussed is the vaccine booster policy in Indonesia. To find out public opinion about this, analyzing sentiment on Youtube video commentary data regarding the vaccine booster policy is necessary. In analyzing sentiment, a suitable classification method is needed so that the polarity of the comment data is in the true polarity class. The raw data resulting from data retrieval from Youtube comments needs to be cleaned, intending to reduce noise in the data, such as deleting junk comments. So in this study, additional cleaning was carried out which was more adapted to the Youtube comment pattern. The classification method used is the Support Vector Machine (SVM) method. SVM is a classification method that has good performance in mapping data according to class and can work well on high-dimensional data. In the SVM method, it is necessary to select a kernel to train the model. In this study, experiments were carried out on both kernels, namely the linear kernel and the RBF. As for the results of the study, the model with the best performance was obtained based on testing with 3929 training data and 364 test data. Accuracy, precision, recall, and f1-score were obtained respectively 0.86, 0.85, 0.82, and 0.83. This model is built with a linear kernel, while the RBF kernel has lower evaluation results regarding accuracy performance and computation time. Then the results of the sentiment analysis regarding the booster vaccine were obtained, namely, from 364 comments, it was predicted that 210 negative comments and 154 positive comments were obtained. It can be concluded that public sentiment, especially Youtube users, tends to reject the vaccine booster policy.
Download
book
BibTex
Latex, Jabref
cloud_download
Original Resource
url resource Institution
