This study analyzes public sentiment toward the naturalization of football players using the K-Nearest Neighbor (KNN) method and the Synthetic Minority Oversampling Technique (SMOTE). KNN is employed for sentiment classification, while SMOTE addresses class imbalance in the dataset. The methodology includes data collection, labeling, cleaning, preprocessing, classification, and model evaluation using Google Colab and Python. The results indicate that without SMOTE, the model performs better, achieving high precision, recall, F1 score, and accuracy. In contrast, applying SMOTE reduces performance, particularly in precision and F1 score. The "Manhattan Neighbor 7" and "Manhattan Neighbor 3" models without SMOTE demonstrate near-perfect results, while SMOTE significantly decreases several evaluation metrics. Additionally, the analysis of public opinions on YouTube reveals a tendency toward negative sentiment in podcasts about player naturalization, hosted by Bung Towel and Anjas Asmara, reflecting public skepticism and critical views on the topic. This study provides valuable insights into public sentiment and the effectiveness of classification methods in the context of national sports issues.