<![CDATA[In the digital era, the internet has become the main source of information for the public, including when searching for hotel information. One platform frequently used for finding hotel information is Google Review, which allows hotel guests to share their experiences. This study aims to analyze sentiments towards hotel reviews, specifically for hotels in Jakarta, through Google Review to help customers choose hotels that meet their needs. The research method uses Serp API in Python to gather review data from Google Review, followed by data preprocessing and labeling with VADER. Sentiment classification is conducted using a stacking ensemble method. The stacking ensemble method in this study employs Naive Bayes, Random Forest, Logistic Regression, Support Vector Machine, K-Nearest Neighbor, and Extra Trees Classifier as base algorithms. The sentiment classification results are then evaluated using precision, F1-score and accuracy metrics. The research findings show that the stacking ensemble method with Naive Bayes, Random Forest, Logistic Regression, Support Vector Machine, K-Nearest Neighbor, and Extra Trees Classifier algorithms performs better than individual models. The precision, F1-Score and accuracy of the stacking ensemble method reach 91%, 97%, 98%, respectively.]]>
Copyrights © 2024
