The problem of class imbalance in sentiment classification is tackled in this research by using webscraped reviews of Montenegrin hotels. The problem is addressed by using support vector machine (SVM) preprocessing and comparing it to the undersampling approach. This study uncovers efficient ways for managing class imbalance, ensuring robust sentiment analysis of hotel reviews in Montenegro's rebuilding tourism industry. Hotel review sentiment analysis is critical for marketing decisions since it translates unstructured feedback into usable insights. It helps organizations to connect their services with the preferences of their customers, resulting in increased customer happiness, loyalty, and revenue growth.