Sentiment Analysis Against Political Figure’s Billboard During Pandemic
Using Naïve Bayes Algorithm

Sentiment Analysis Against Political Figure’s Billboard During Pandemic
Using Naïve Bayes Algorithm

covid-19 pandemic; twitter; sentiment analysis; naïve bayes clalssification algorithm

In the midst of the Covid-19 Pandemic, many Indonesians have reacted negatively to the placement of political individuals'
billboards with very huge sizes on the streets. The early political campaign that was run was thought to be contentious. On
social media like Twitter, the majority of people freely share their thoughts. The purpose of this study is to investigate how the
general public reacted to the placement of billboards advertising political figures during the epidemic and to categorize those
responses. It is envisaged that it would also provide advice for connected parties that may be used when making judgments
regarding the policy of constructing billboards for political figures during a pandemic based on the results of data analysis.
Twitter users tend to be more expressive because of the character limits, which means they have sentimental or emotional
values. Using the Nave Bayes Algorithm, it is possible to do sentiment analysis on the sentiment data by categorizing user
comments into positive, negative, and neutral attitudes. Regarding the sentiments expressed on billboards showing political
leaders during the pandemic, tweets were sorted into three categories: liked, unfavorable, and neutral. The accuracy rate from
Naive Bayes categorization of political personalities during the pandemic on social media Twitter was 83.3% with a precision
value of 89%, recall 83%, and f-1 score of 82%.

Ade Bastian1
, Ardi Mardiana2
, Dinda Sri Wulansari3

Universitas Majalengka

03-02-2023

Fajar bagus W

PDF

Indonesia

Text