Abstract:
Sentiment analysis is an essential task in natural language processing that is critical to
comprehending user attitudes and public opinion in a variety of fields. This study compares
and thoroughly examines several machine-learning techniques for sentiment analysis.
Linear regression, Decision Tree, Random Forest, XGBoost, KNN, Artificial Neural
Network (ANN), and Convolutional Neural Network (CNN) are among the models that
were examined. The main metric used in the study to assess each model's performance is accuracy. With an accuracy of 0.9823%, (TPOT) was only behind Linear Regression, which had an
amazing accuracy of 0.9825%. Moreover, Decision Tree and Random Forest performed
admirably, with respective accuracies of 0.9762% and 0.9805%. On the other hand, the
accuracy obtained by XGBoost, KNN, and ANN were 0.9693%, 0.9753%, and 0.9783%,
in that order. Remarkably, the convolutional neural network (CNN) demonstrated a significantly
reduced accuracy of 0.8199%, suggesting possible difficulties when utilizing this
architecture for sentiment analysis inside the specified framework. The research's conclusions provide important new information on which machine learning models are best suited for tasks involving sentiment analysis. Based on the particular needs of their applications, researchers and practitioners can use these data to help them choose a sentiment analysis model. The study also emphasizes how crucial it is to take into account a variety of machine learning techniques to improve the precision and dependability of sentiment analysis systems in practical contexts.
