<![CDATA[This study compares three popular text classification algorithms—Naive Bayes (NB), Support Vector Machine (SVM), and k-Nearest Neighbour (k-NN)—for classifying Spotify review data. The significance of this topic lies in applying these algorithms to sentiment analysis, which can help better understand user feedback. The research method involves testing these algorithms on Spotify review data classified into positive and negative categories. Results show that the k-NN algorithm achieves the highest accuracy at 83.67%, while NB and SVM achieve accuracies of 77.67% and 76.50%, respectively. The AUC values are 0.950 for NB, 0.955 for SVM, and 0.914 for k-NN. Despite k-NN showing the highest accuracy, SVM exhibits the highest AUC, indicating very good performance in distinguishing between categories. In conclusion, while k-NN demonstrates superior accuracy, a comprehensive evaluation based on various metrics is crucial for selecting the optimal algorithm for sentiment analysis]]>
Copyrights © 2024
