Analisis dan Visualisasi Sentimen Analisis Data Twitter Menggunakan Support Vector Machine dan Social Network Analysis

Abstract

The dissemination of information in today's era is extremely rapid due to the development of various social media platforms. However, the information circulating is sometimes inaccurate or invalid. One of the platforms commonly used to spread information is Twitter. Among the vast amount of circulating information, some content is deliberately spread—either positive or negative. This sparked the idea to develop a sentiment analysis program that can be visualized using Support Vector Machine (SVM) and Social Network Analysis (SNA). The model was built using a Twitter dataset and achieved an accuracy of 86.8%, a precision of 86.2%, a recall of 87%, and an F1-score of 86.6%. The model effectively classifies sentiment in Twitter data into appropriate categories. The visualization is presented by displaying nodes and edges interconnected to form a graph. Each node varies in size and color to represent the relationships found in the data.