<![CDATA[The high number of cases of HIV/AIDS infection which continues to increase, especially among young people or adolescents, is a serious problem. HIV data on children in Indonesia from 2020 to September 2022, there are 12,553 children aged 14 years and under who are infected with HIV with boys predominate. This shows that efforts to prevent HIV transmission from mother to child need to be increased. As social media develops, people often talk about and give their opinions through various media, one of which is social media Twitter. Opinions given by the community regarding HIV/AIDS cases varied, such as HIV screening, mandatory use of condoms so as not to get sexually transmitted diseases, and HIV testing before marriage. This research is expected to be useful in helping to conduct research on public opinion that contains sentiments, positive sentiments and negative sentiments. The method used in this study is Naive Bayes, with the process of crawling data from Twitter using Rapid Miner, preprocessing data that has been obtained from Twitter using cleansing, tokenization and filtering. For the classification process using the Naive Bayes method. The data used are tweets in Indonesian with the keywords HIV and AIDS, with a total dataset of 3.051 tweets. After going through the data selection process, the final data was 1.134 tweets with a positive sentiment of 63.49% and a negative sentiment of 36.51%. The results of testing the accuracy performance of the naïve Bayes method using RapidMiner tools obtained the best average accuracy value in the first test scenario with an average accuracy value of 59.53%, a Precission value of 68.68% and a recall value obtained of 64.89%. While the results of the accuracy performance test on Python tools using the Naïve Bayes method obtained an accuracy value of 66.00%, a precision value of 67.45% and a recall value of 94.07%]]>
