ANALISIS SENTIMEN PENGGUNA TWITTER TERHADAP KENAIKAN HARGA BBM DI INDONESIA DENGAN ALGORITMA NAIVE BAYES
Total View This Week0
Total View This Week0
Institusion
Universitas Nahdlatul Ulama Surabaya
Author
MARVYNA, HANIFAH CITRA
Subject
T Technology (General)
Datestamp
2023-03-29 03:45:56
Abstract :
The difficulty of manually classifying text data makes sentiment analysis a solution to make it easier to classify data polarity. Twitter is one of the social media that provides concise text data in the form of opinions on a topic. One of the topics that is currently being discussed is the increase in the price of fuel oil (BBM) in Indonesia. Through this research, an analysis of the sentiments of Twitter users towards rising fuel prices in Indonesia is carried out using the Naive Bayes algorithm. The method used in this study starts from data collection, text preprocessing, data labeling, feature extraction, data s, classification and evaluation of algorithm performance. The results of the analysis of Twitter users' sentiment analysis on rising fuel prices in Indonesia with the Naïve Bayes Algorithm are dominated by negative sentiments from 5,000 collected tweet data divided into 54.6% negative sentiment, 31.8% positive sentiment, and 13.6% neutral sentiment . While the best results of the classification performance of the Naïve Bayes algorithm were obtained in the 80:20 data ratio experiment with an accuracy value of 65%, a precision of 74%, a recall of 45%, and an error rate of 35%.
The difficulty of manually classifying text data makes sentiment analysis a solution to make it easier to classify data polarity. Twitter is one of the social media that provides concise text data in the form of opinions on a topic. One of the topics that is currently being discussed is the increase in the price of fuel oil (BBM) in Indonesia. Through this research, an analysis of the sentiments of Twitter users towards rising fuel prices in Indonesia is carried out using the Naive Bayes algorithm. The method used in this study starts from data collection, text preprocessing, data labeling, feature extraction, data s, classification and evaluation of algorithm performance. The results of the analysis of Twitter users' sentiment analysis on rising fuel prices in Indonesia with the Naïve Bayes Algorithm are dominated by negative sentiments from 5,000 collected tweet data divided into 54.6% negative sentiment, 31.8% positive sentiment, and 13.6% neutral sentiment . While the best results of the classification performance of the Naïve Bayes algorithm were obtained in the 80:20 data ratio experiment with an accuracy value of 65%, a precision of 74%, a recall of 45%, and an error rate of 35%.
Download
book
BibTex
Latex, Jabref
cloud_download
Original Resource
url resource Institution
