<![CDATA[The study aims to investigate the effectiveness of sentiment analysis algorithms, specifically Support Vector Machine (SVM) and Decision Tree (DT), integrated with the Synthetic Minority Over-sampling Technique (SMOTE) to mitigate class imbalance issues. Guided by the Cross-Industry Standard Process for Data Mining (CRISP-DM) framework, the research involves several stages: Business Understanding, Data Understanding, Data Preparation, Modeling, Evaluation, and Deployment. The process begins with understanding the business objectives of sentiment analysis and proceeds to explore and prepare the dataset for analysis. SVM and DT algorithms, enhanced with SMOTE, are then implemented for sentiment classification. The study reveals promising results in sentiment analysis tasks. When integrated with SMOTE, SVM achieves an accuracy of 99.21%, while DT attains an accuracy of 98.33%. The Area Under the Curve (AUC) metrics indicate high confidence in classifying positive instances, with SVM and DT demonstrating AUC scores of 1.000 and 0.996, respectively. These findings underscore the efficacy of SVM and DT algorithms, enhanced with SMOTE, in accurately classifying sentiment within text data, thereby addressing class imbalance issues effectively]]>
Copyrights © 2024
