Sentimental analysis is use for analyzing the text so that it can determine theemotions what is in the message which can be positive, negative, or neutral. Social media platforms such as twitter, Instagram, Facebook, etc. where public opinions have been expressed a lot. Other sentimental analysis algorithms include the rules-based and hybrid approach for data processing. Sentiment algorithms are also analysed accordingly with in forms of text like sentences, It also contains relevantinterpretations through the information that was provided. In this study, a baseline model was first established, achieving an accuracy to serve as a comparison point for more sophisticated models. The Logistic Regression model outperformed the baseline significantly, demonstrating its effectiveness in accurately classifying sentiments. The Decision Tree Classification model, while an improvement over the baseline was less accurate than Logistic Regression, suggesting potential issues with overfitting and data dependency. The Random Forest Classification model provided a robust alternative, matching the performance of Logistic Regression and benefiting from the ensemble approach to handle diverse patterns in the sentiment data. Keywords: Sentimental analysis, Machine learning, supervised learning, twitter
Read full abstract