<![CDATA[The Covid-19 epidemic has caused substantial changes in Indonesia, causing the government to create SATUSEHAT Mobile, formerly known as PEDULILINDUNGI, which will be formally renamed on March 1, 2023. This software tracks the spread of Covid-19, offers vaccination information, and distinguishes distinct zones. Users can share their location data for travel purposes, allowing Covid-19 patients to be contacted. Sentiment analysis is used to assess SATUSEHAT users' perspectives based on Play Store reviews, with an emphasis on positive and negative comments. Textual data connected to certain items or entities is analyzed and modified using Data Mining techniques. The SATUSEHAT application was tested in this study. The program classified 43 positive and 1662 negative comments correctly, yielding 1705 successful classifications out of 1861 comments. The data from the confusion matrix allowed for the calculation of accuracy, precision, and recall, achieving 92% accuracy, 93% average precision, and 98% average recall. According to the research findings, the Naive Bayes algorithm with TF-IDF Vectorizer is the best at producing positive and negative labels with 92% accuracy, even for unbalanced data. In comparison to other algorithms, Naive Bayes with TF-IDF Vectorizer exhibited good accuracy, indicating a promising topic for further study.]]>
Copyrights © 2023
