A Comparative Sentiment Analysis about HIV and AIDS on Twitter Tweets Using Supervised Machine Learning Approach

Sentiment analysis plays a crucial role in understanding public attitudes towards important social and political issues discussed on social media platforms. This study focuses on analyzing sentiment in Twitter data related to HIV and AIDS, employing a supervised machine learning approach. The primary objective is to compare the effectiveness of various machine learning algorithms in categorizing the sentiment of tweets. The researcher collected and pre-processed a dataset of tweets using sentiment analysis methods from TextBlob and Vader. Three classification algorithms, namely multinomial Naive Bayes, support vector machine, and logistic regression, were evaluated for their performance in sentiment classification. Our findings demonstrate that support vector machine coupled with an n-gram model achieved the highest accuracy, reaching 99% in sentiment classification. This comparative analysis makes a novel contribution to the field of sentiment analysis, specifically in the context of Twitter data related to HIV and AIDS. The insights gained from this study provide valuable implications for understanding public attitudes towards HIV and AIDS on social media and can inform the development of effective public health messaging and policy initiatives pertaining to this critical topic.