<![CDATA[The landscape of sentiment analysis applications in Indonesia is on the rise with the many published papers on the subject over the years. The need to predict sentiment coincides with the rise of social media and how the public uses it to express sentiments toward an interesting topic. The lack of tools for working with the Indonesian language has brought the invention of libraries to tackle the difficulty and uniqueness of the language on various topics from diverse data sources. The introduction of Sarkawi as a stemmer helps researchers overcome dimensionality problems commonly found with text processing, and boosts the performance of machine learning (ML) models. Using InSet as a lexicon dictionary capable of performing sentiment prediction has started gaining popularity for automatic labeling. The development of IndoBERT, an advanced neural network (NN) large language model (LLM) specifically trained from a large Indonesian text corpus capable of more than sentiment analysis, has gained traction both for automatic labeling and prediction models. Although the majority of research revolves around Naïve Bayes (NB), State Vector Machine (SVM), and K-Nearest Neighbor (KNN) the future of sentiment analysis applications in Indonesia could be heading towards a more advanced deep learning architecture. Finally, this study is intended as a basis for future research in the applications of sentiment analysis in Indonesia and the development of the language.]]>
Copyrights © 2024
