Various moments are immortalized in social networking, one of them Twitter. Twitter is one of the social media that allows users to interact, share information, or even to express feelings and opinions as well as in expressing opinions about "Mudik". Every tweets not all contain the same value, therefore there needs to be a sentiment analysis and tweet classification that discusses about to know the amount of public sentiment to mudik year 2019. This will be a concern for “facilicies transportation for goinghome” next year. One of method used is classification method using k-NN, SVM, Naïve Bayes and Neural Network. The results of this study can be seen that positive sentiments appear more than the value of other sentiments for Mudik 2019, next time negative’s opinion appear more than positive Mudik 2019. and k-NN algorithm has a higher accuracy, it is accuracy=90.76 and AUC=0.939, SVM generated accuracy=89.03% and AUC=0.500, Naïve Bayes generated accuracy=78.16% and AUC=0.567 then Neural Network generated accuracy=52.73% and AUC=0.000.
Copyrights © 2019