期刊名称：Journal of Theoretical and Applied Information Technology
印刷版ISSN：1992-8645
电子版ISSN：1817-3195
出版年度：2013
卷号：58
期号：1
出版社：Journal of Theoretical and Applied
摘要：The national exam (UN) is a government policy to evaluate the education level on a national scale to measure the competence of students who graduate with those of other schools at the same educational level. The policy in conducting the UN is always a topic of discussion and phenomenon that is covered in various media, because it causes various problems that become pro and contra in society. An analysis sentiment or opinion mining in this research is applied to analyze public sentiment and group polarities of opinions or texts in a document, whether it shows positive or negative sentiment, in conducting the UN. The analytical process and data processing for document classification uses two classification methods: the quintuple method and one of the methods for learning machines, the Naive Bayes Classifier (NBC) method. The data used to classify documents is in the form of news text documents about conducting the UN, which is taken from online news media (detik.com). The data gathered comes from 420 news documents about conducting the UN from 2012 and 2013. Based on the analytical results and document classifications, it has been found that public sentiment towards carrying out the UN in 2012 and 2013 shows negative sentiment. The results of the data processing and document classification of conducting the UN overall show a positive opinion of 32% and a negative opinion of 68%. The results of the document classification are based on the polarization of public opinion about carrying out the UN, which reveal in the opinion category of carrying out the UN in 2012 that there was a positive opinion of 44% and a negative opinion of 56%. Meanwhile, for conducting the UN in 2013, there is a positive opinion of 20% and a negative opinion of 80%. These results reveal an increase in negative public sentiment in conducting the UN in 2013.
关键词：Sentiment Analysis; Opining Mining; National Exam; Quintuple; Naive Bayes Classifier
