Analisis Sentimen Masyarakat Kalimantan Tengah Terhadap Perkebunan Kelapa Sawit Menggunakan TF-IDF dan Support Vector Machine

Abstract

The transformation towards Society 5.0 has had a significant impact on the rapid growth of data available worldwide, both useful and less directly beneficial, known as big data. This phenomenon provides opportunities for researchers to leverage big data as a valuable source of information, provided it is processed and analyzed using appropriate methods. One of the rapidly growing applications is sentiment analysis, which extracts insights from text data, such as that gathered from social media platforms. This study applies the TF-IDF feature extraction technique and the SVM (Support Vector Machine) classification method to perform sentiment analysis on Twitter text data. The results of the research show that the model built using the combination of TF-IDF and SVM achieved an accuracy of 86%, with precision, recall, and F1-Score values of 85% each. These findings indicate that the application of TF-IDF with SVM provides optimal performance in sentiment analysis, considering the word frequency within documents, and makes a significant contribution to processing big data for more accurate and effective sentiment analysis