Farsyah, Adella Fatikah (2026) Analisis Sentimen Terhadap Akurasi Opini Publik dengan Tagar #Kaburajadulu Menggunakan Algoritma Machine Learning. Other thesis, Universitas Islam Riau.

Text skripsi_213510173_watermark.pdf - Published Version Restricted to Registered users only Download (2MB) | Request a copy

Abstract

This study analyzes public sentiment toward the hashtag #KaburAjaDulu based on 2,003 tweets collected through web scraping using the tweet-harvest tool. Public opinion is categorized into 560 positive sentiments and 1,298 negative sentiments, and then analyzed using three machine learning algorithms: Naive Bayes, Support Vector Machine, and K-Nearest Neighbors. This study aims to analyze public opinion regarding the social issues reflected in the hashtag #KaburAjaDulu and to determine the best algorithm for sentiment analysis. The performance of the three algorithms is compared using data split ratios of 60:40, 70:30, 80:20, and 90:10 to measure accuracy. The data are processed through cleaning, case folding, normalization, tokenization, stopwords removal, and stemming, and the SMOTE technique is applied to address data imbalance. The results show that K-Nearest Neighbors has a fairly good ability to predict sentiment; however, in highdimensional text data, KNN becomes less effective, resulting in lower accuracy compared to SVM. Based on the experimental results, Support Vector Machine achieves the best accuracy of 91% with a 90:10 data split ratio, followed by KNearest Neighbors with an accuracy of 82%, and Naive Bayes with an accuracy of 77% after applying SMOTE at the same ratio. These findings indicate that Support Vector Machine is the best algorithm for sentiment analysis of tweets containing the hashtag #KaburAjaDulu.

Actions (login required)

View Item