<![CDATA[Traffic congestion occurs in many places throughout Indonesia, especially in its capital region of Jakarta. Many strategies have been executed by the government of the capital region as a mean to solve the ongoing traffic congestion problem, one of them is the 'odd-even' policy. On the other note, the problem has inflicted a wide social media complains among Jakarta's residents. In this case, Twitter is considered as a relatively fast and effective social media platform to post opinions used by many Indonesians. Considering its large number of users and easy access to public's opinions, Twitter will have a lot of public's opinions' data which can be used as a material to evaluate the 'odd-even' policy in the capital region of Jakarta. Therefore a method which can separate sentiment from user is needed. It's to answer whether the sentiment is categorized as positive or negative class. In this study, the researcher used BM25 method and K-Nearest Neighbor (KNN) as classifiers. The best test results for f-measure values are 66,1% while the results of accuracy is 66,5%.]]>
Copyrights © 2019
