<![CDATA[This study was conducted with the aim of analyzing the sentiments contained in public reviews of the Gelora Sriwijaya International Stadium in Palembang. The analysis utilized the Random Forest Classifier algorithm as the main method for the classification process. The data source was obtained from Google Maps through web scraping using the Instant Data Scraper tool, resulting in a total of 1,000 reviews written between 2019 and 2025. These reviews then went through a series of preprocessing stages, including data cleaning to remove irrelevant information, case folding to standardize letter formats, stemming to return words to their root form, stopword removal to eliminate common words with little semantic value, and tokenization to split the text into individual word units. Subsequently, the text data was transformed into a numerical representation using the Term Frequency–Inverse Document Frequency (TF-IDF) method to make it suitable for processing by the classification algorithm. Sentiment classification was carried out by categorizing reviews into two main classes: positive sentiment and negative sentiment. The evaluation results showed that the model achieved an accuracy of 89% when tested using the K-Fold Cross Validation technique, which was higher than the 84% accuracy obtained without validation. These findings demonstrate that the Random Forest Classifier algorithm is highly capable of performing sentiment analysis and has the potential to serve as a useful tool for stadium management to understand public perceptions and improve the quality of facility management.]]>
Copyrights © 2025
