Implementasi Algoritma Random Forest dan Model Bag of Words Dalam Analisis Sentimen Mengenai E-Materai

Abstract

For CPNS registration in 2024, a digital stamp purchase method will be implemented via the website https://meterai-elektronik.com as a requirement for complete administration. However, the page that provides e-stamps experienced an error because many visitors entered and tried to buy e-stamps at the same time. This disruption has resulted in several people having opinions about e-stamps. To find out people's sentiments about e-stamps after this incident, sentiment analysis was carried out. In analyzing these sentiments, the Random Forest algorithm is used with the Bag of Words model, where Random Forest is a development method of Classification and Regression Trees (CART) which is said to be more precise in predicting and the Bag of Words model is widely used with good results for predicting language modeling and classification. documents because Bag of Words is simple and flexible. In sentiment analysis, 86% of neutral opinions were produced, 9% of negative opinions and 5% of positive opinions. Then the Random Forest algorithm with the Bag of Words model got an accuracy value of 70.1%, precision of 50.7%, recall of 54.1%, and F1-Score of 47.6%.