Perbandingan Sentimen Analisis Tweet Berbahasa Indonesia Menggunakan Metode Naive Bayes dan Support Vector Machine
Total View This Week7
Total View This Week7
Institusion
STMIK Bumi Gora
Author
Rialdi, Lalu Ivan
Subject
T Technology (General)
Datestamp
2021-09-06 05:51:24
Abstract :
Twitter has been used extensively by all levels of society in recent years especially in the country of Indonesia. Prior to the general media election of twitter, it is a place to aspire to political or public leaders. In the elections there are certain parties or institutions who would like to know opinions and responses to political figures. The politically assessed figure is the one who is considered worthy and has the ability to be elected leader. Many studies involving tweets for public or political figures. The analysis is done with the rating of dataset tweets from twitter users of certain political figures. Some methods are to be made clear of the data first Naive Bayes and the two vector-machine support and combined with a neutering feature using tf-idf. The rating on twitter's dataset is positive, negative and neutral made up of subjects, precision, recall andf1-score. Testing on the application and the notebook's tool jupyter showed that classified by tf-idf's tf-idf method of the naive's file was not bad, but it was still under the claim that vector-machine was nearing its perfect classification machine with a 0.98 accuracy score. Keyword : Sentiment analysis, Twitter, Naive Bayes, Classifier(NBC), Support Vector Machine(SVM)
Twitter has been used extensively by all levels of society in recent years especially in the country of Indonesia. Prior to the general media election of twitter, it is a place to aspire to political or public leaders. In the elections there are certain parties or institutions who would like to know opinions and responses to political figures. The politically assessed figure is the one who is considered worthy and has the ability to be elected leader. Many studies involving tweets for public or political figures. The analysis is done with the rating of dataset tweets from twitter users of certain political figures. Some methods are to be made clear of the data first Naive Bayes and the two vector-machine support and combined with a neutering feature using tf-idf. The rating on twitter's dataset is positive, negative and neutral made up of subjects, precision, recall andf1-score. Testing on the application and the notebook's tool jupyter showed that classified by tf-idf's tf-idf method of the naive's file was not bad, but it was still under the claim that vector-machine was nearing its perfect classification machine with a 0.98 accuracy score. Keyword : Sentiment analysis, Twitter, Naive Bayes, Classifier(NBC), Support Vector Machine(SVM)
Download
book
BibTex
Latex, Jabref
cloud_download
Original Resource
url resource Institution
