Optimization of Classification Algorithm with GridSearchCV and Hyperparameter Tuning for Sentiment Analysis of the Nusantara Capital City

The relocation of the country's capital has implications for social, economic, geographical and political aspects. This raises the public's views which are expressed in Twitter tweets that contain public sentiment. Sentiment analysis has a significant influence on understanding public views. Therefore, further research is needed related to the sentiment of analysis of the IKN. In this study, a model that can predict public sentiment is created by developing a model at the training stage using the GridSearchCV algorithm and hyperparameter tuning with several classification algorithms. The best model was produced with the Support Vector Machine classification algorithm that was able to outperform, compared to probability-based and tree-based classification algorithms with an accuracy of 69.95%.