This article use Systematic Literature Review (SLR) to classify sentiment analysis based on case studies, methods, social media, and platforms. The coding stage is divided into three stages, namely the open, selective and axial coding. The literature study on sentiment analysis is divided into two parts: identifying gaps based on case studies and data sources and identifying gaps based on the methods or algorithms used. The gap identification results based on case studies and data sources show that popular review topics are synonymous with entertainment, economic and political content. Therefore, the quantity of research with review topics related to the implementation of education, the dynamics of the bureaucracy, health facilities and services, and non-governmental organization’s activities need to be increased. Meanwhile, the most dominant platforms used as data sources are website and mobile-based applications. The results of the gap identification based on the method and algorithm show that the quantity of research with the Naive Bayes Classifier (NBC) and Support Vector Machine (SVM) method or algorithm is more dominant than the k-Nearest Neighbor (k-NN) and Lexicon-based. Thus, it is necessary to increase the number of other classification methods such as Particle Swarm Optimization, BM25, Decision Tree, K-Means, and Neural Networks.