<![CDATA[The Indonesian government made a policy to stop consuming syrup as a form of prevention against acute kidney failure, which affects many people in Indonesia. However, the policy has caused a lot of comments from the public. These public comments can be found on YouTube, because YouTube has a large data source opportunity to be used as a research material. These comments can be processed directly without using a machine, but it is less effective and efficient. Thus, the comments are processed using machine learning methods. Based on the earlier research, the naive bayes classifier algorithm tends to be simple and easy to use. In addition, this algorithm also has a high accuracy. The amount of data used in this study is 1000 YouTube comment data related to videos regarding the policy of prohibiting the use of syrup medicine, the comments are divided into 2 category, which are positive class and negative class. The results of labeling 1000 comments obtained 704 negative comments and 296 positive comments. Based on the experiments conducted using python programming language, the highest accuracy was obtained at 74% in 70:30 data split. Furthermore, in the balanced dataset (296 positive and 296 negative comments), the highest accuracy was obtained at 64.70% with in 80:20 data split. These results represent that the naive bayes classifier algorithm is good enough at sentiment analysis about the policy of prohibiting the use of syrup drugs.]]>
Copyrights © 2023
