Changes Analysis in Public Opinion Regarding Binary Option Trends using K-Means

• Main Authors: Maylawati, Dian Sa'Adillah, Abdullah, Rifqi Fajri, Gerhana, Yana Aditia, Wahana, Agung, Budiman, Ichsan, Uriawan, Wisnu
• Format: Conference Proceeding
• Language: English
• Subjects: binary option; Blogs; clustering; Clustering algorithms; Elbow; Government; Investment; k-means; machine learning; opinion analysis; Social networking (online)
• ISSN: 2767-7826
• DOI: 10.1109/ICCED60214.2023.10425589

Binary option-based applications are being discussed in Indonesia because several influencers are promoting the application. But the government said the app is included in gambling, which makes the app illegal. Some influencers have become associated with binary options cases, which have made people busy giving opinions about binary options on social media. This study aims to analyze the changes in public opinion regarding binary options in Indonesia. This research use K-Means++ algorithm to construct the opinion clusters. The data used in this study was taken from Twitter social media from January 1,2022, to March 31,2022, with total 19,669 data that divided into two moments, between before and after the case of binary options in Indonesia. Based on result of Elbow method, there are 4 and 6 optimum cluster that can be analyzed. Clustering with the K-Means++ algorithm gets good results with an average of silhouette coefficient value of 0.6605. The opinion analysis result show that before the case of the arrest of binary option application affiliates in Indonesia, people tended to express neutral opinions. However, after the case was widely discussed and confirmed as fraudulent investment scams, Indonesian people still expressed opinions in neutral language but agreed that binary options applications are not trading and illegal under Indonesian law.