<![CDATA[Advancements in the field of Natural Language Processing (NLP) have opened significant opportunities in sentiment analysis, particularly in the context of disaster response. In today's digital era, YouTube has emerged as a primary source for the public to acquire information regarding critical events. This study explores and compares two dominant sentiment analysis techniques, namely Naive Bayes and Support Vector Machine (SVM). It utilizes YouTube comment data related to natural disasters to test the effectiveness of these algorithms in identifying and classifying public sentiment as neutral, positive, or negative. The process involves collecting comment data, pre-processing the data, and applying Term-Frequency-Inverse Document Frequency (TF-IDF) weighting to prepare the data for analysis. Subsequently, the performance of both models is evaluated based on metrics such as accuracy, precision, recall, and F1 score. The results indicate that while both algorithms have their strengths and weaknesses, SVM tends to show better performance in sentiment classification, especially in terms of accuracy and precision, with an accuracy result of 92% and precision of 89% for negative predictions and 94% for positive predictions. On the other hand, Naive Bayes only achieved an accuracy of 79% and a precision of 91% for negative predictions and 73% for positive predictions. This study provides significant insights into the application of machine learning algorithms in sentiment analysis.]]>
Copyrights © 2024
