JURNAL TEKNIK INFORMATIKA DAN KOMPUTER
Vol. 2 No. 1 (2023): Jurnal Teknik Informatika dan Komputer

Perbandingan Akurasi Metode Naïve Bayes Classifier dan Lexicon Based Pada Analisis Sentimen Respon Masyarakat Tentang Kebijakan Kenaikan Harga Minyak Goreng

Faldy Irwiensyah (Universitas Muhammadiyah Prof. Dr. Hamka)
Firman Noor Hasan (Universitas Muhammadiyah Prof. Dr. Hamka)



Article Info

Publish Date
29 May 2023

Abstract

Cooking oil is a basic need for Indonesian people. Indonesia experienced a shortage of oil in March 2022. This was a hot topic of discussion on social media Twitter last March, many people thought positively or negatively. However behind it all, there are differences in the assessment of parties who feel the pros and cons, various parties have different points of view. In this article, we conduct a sentiment analysis of public responses regarding the scarcity of cooking oil using a dataset obtained from the Twitter digital platform. This article aims to classify tweets related to the scarcity of cooking oil into positive and negative sentiments using a machine learning strategy with the Naive  Bayes and lexicon based methods. This algorithm was chosen to make it easier for interested users to compare methods and find out how accurate it is, which is where the level of accuracy obtained from the lexicon method is 42% and the method using the naïve Bayes classifier is 72%. Shows the results of the analysis related to the scarcity of cooking oil for the highest level of accuracy, namely the method that uses the naïve Bayes classifier compared to the method that uses lexicon based

Copyrights © 2023






Journal Info

Abbrev

jutikom

Publisher

Subject

Computer Science & IT

Description

Jurnal Teknik Informatika dan Komputer (JUTIKOM) is a national forum for scientists and engineers involved in all aspects of informatic engineering and computer science to publish high quality and refereed papers. Topics covered by Jurnal Teknik Informatika dan Komputer include but are not limited ...