Safitri, Sandra (2025) Analisis Sentimen Pengguna Twitter Terhadap Kinerja Pemerintahan Prabowo Setelah 100 Hari Menjabat dengan Metode Random Forest. Other thesis, Universitas Islam Riau.

Text skripsi_183510648_watermark.pdf - Published Version Restricted to Registered users only Download (2MB) | Request a copy

Abstract

Social media has become the main means for people to express their opinions on government policies openly. Twitter, with more than 24 million active users in Indonesia, is often used to respond to public issues in real-time. To understand the public sentiment towards the performance of Prabowo Subianto's administration as the new President of Indonesia, it is necessary to analyze the opinions spread on social media, especially Twitter. However, the main challenge in this analysis is the diversity of expressions and the complexity of the language used by users. Therefore, Machine Learning-based approaches are used to automate the sentiment classification process more accurately and efficiently. This research aims to build and evaluate a sentiment classification model using the Random Forest algorithm on Indonesian tweet data related to government performance in the first 100 days of office. The methods used include data crawling, preprocessing, semi-automatic labeling, feature extraction with TF-IDF, and model training with various parameter variations. The results showed that the best model was obtained at a configuration of 200 trees and a training-test data ratio of 90:10 with an accuracy of 86.9% and an F1-score of 88%. Compared to increasing the number of trees, increasing the proportion of training data has a greater impact on improving classification performance.

Actions (login required)

View Item