Open AccessSentiment analysis of COVID-19 vaccine in Indonesia using Naïve Bayes Algorithm
Author(s) -
Pristiyono,
Mulkan Ritonga,
Muhammad Ali Al Ihsan,
Agus Anjar,
Fauziah Hanum Rambe
Publication year - 2021
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/1088/1/012045
Subject(s) - crawling , indonesian , naive bayes classifier , covid-19 , computer science , government (linguistics) , social media , sentiment analysis , medicine , world wide web , artificial intelligence , support vector machine , pathology , disease , philosophy , linguistics , infectious disease (medical specialty) , anatomy
As of January 2021, with 2,066,175 deaths, 95,612,831 confirmed cases have been reported globally. Indonesia’s COVID-19 Task Force report shows that there are currently 27,203 deaths, with reported cases exceeding 951,651, among the highest in Asia. The President of the Republic of Indonesia created a national team to speed up the production of vaccines for COVID-19. It stipulates that the government will arrange the provision, delivery, and vaccination of COVID-19 vaccines. The vaccination scheme would then become the pros and cons of Indonesian society. This research assesses the opinion of the Indonesian people through a social network analysis of the COVID-19 vaccine in January 2021. We used sentiment analysis using Naïve Bayes Algorithm by crawling Twitter data with ‘Vaccine COVID-19’ as keywords. We perform the data crawling process manually using the access token received from the Twitter API using the Rapid miner tools to extract the requested information and data. Data crawling continued with the Drone Emprit Academic Streaming Public Twitter Tool because of limited manual crawling resulting in more than 6000 tweets related to selected keywords on January 15-22, 2021. The result of sentiment measurement with over 3.4 thousand negative tweets (56%), over 2.4 thousand positive tweets (39%), and the remaining 301 tweets (1%) was neutral during the period of the week.