Naïveve Bayes and Logistic Regression for sentiment analysis and emotion detection from text

This paper presents a comprehensive approach to emotion detection using machine learning by transforming text data into numerical vectors. It standardizes datasets into “Text” and “Label” columns, converting emotions to binary classification. Naïveve Bayes and Logistic Regression were selected, with hyperparameter tuning conducted via GridSearchCV and 5 -fold cross-validation. Naïveve Bayes used alpha and fit_prior parameters, while Logistic Regression used C and solver. Logistic Regression showed robust performance, highlighting the importance of data preprocessing and hyperparameter tuning in enhancing model accuracy.