Analisis Sentimen Kendaraan Listrik Pada Media Sosial Twitter Menggunakan Algoritma Logistic Regression dan Principal Component Analysis

Youga Pratama (Telkom University, Bandung)
Danang Triantoro Murdiansyah (Telkom University, Bandung)
Kemas Muslim Lhaksmana (Telkom University, Bandung)

JURNAL MEDIA INFORMATIKA BUDIDARMA
Vol 7, No 1 (2023): Januari 2023

Twitter sentiment analysis is a method for identifying a person's opinions, reactions, judgments, evaluations, and emotions towards certain topics on Twitter social media. Opinions or can be called opinions can be classified as positive or negative. This research was conducted to find out public opinion about electric vehicles on Twitter social media, which is more positive or negative. The data obtained was 1874 tweets with data divided into training data and testing data at a ratio of 80:20. Data is classified using the Logistic Regression (LR) method, and Principal Component Analysis (PCA) as an optimization to improve accuracy. In this study it was found that around 86.9% of the opinions were positive, and 13.1% of the opinions were negative on the topic of electric vehicles. The results of research conducted using the Logistic Regression algorithm obtained the best accuracy of 87.9%, and after being optimized using Principal Component Analysis the best accuracy obtained increased to 90%.