| Penulis/Author |
Dr. Sigit Priyanta, S.Si., M.Kom. (1); Prof. Dra. Sri Hartati, M.Sc., Ph.D. (2); Prof. Drs. Agus Harjoko, M.Sc., Ph.D. (3); Prof. Drs. Retantyo Wardoyo, M.Sc., Ph.D. (4) |
| Abstrak/Abstract |
The increasing utilization of Internet has increased the number and types of content on the Internet, particularly text, that became very large and spread out in many sources of information such as blogs, news sites, forums, social networks, and
micro-blogging. It affects the information overload for users. Information overload can be overcome, among others, by a text classification. Therefore, it is necessary to find a system that can identify opinions, attitudes, and sentiments in a text automatically. This study compared subjective and objective sentences classification methods in Indonesian news. The methods compared were rule-based classifier, Naïve Bayes Classifier (NBC), and Support Vector Machine (SVM) classifier. The results of the study examined in 1050 sentences manually labeled as subjective or objective sentences show the accuracy of 80.4%, 74% and 71% for the rule-based classifier, SVM classifier, and NBC, respectively. |
