Analisis Text Mining pada Sosial Media Twitter Menggunakan Metode Support Vector Machine (SVM) dan Social Network Analysis (SNA)

Abstract

Online loans are growing rapidly in Indonesia in the last two years. This is because the online loan administration requirements are easier compared to bank financial service loans. Online loans are financial services that provide online-based services. Along with the development of online loans, many illegal online loans have sprung up and often commit violations, such as leaking customer personal information and abusing data by carrying out extreme actions such as terrorizing customers who make online loan transactions. This certainly gets a lot of comments from the public, especially on social media twitter. This study aims to conduct a sentiment analysis to see what phenomena are happening among the public regarding online loans. The data used are tweets or retweets from Twitter social media with #pinjamanonline #pinjol. Twitter social media was chosen because an incident can become a phenomenon if it gets a lot of attention from the community, especially on Twitter social media. In this study, using text mining techniques by applying the Support Vector Machine algorithm to classify sentiments on twitter users regarding online loans. This study also looks at the interactions that occur on social media Twitter using social network analysis (SNA). the results of research and testing of the Support Vector Machine method to classify online loans with an Accuracy value level of 86.6%, with a positive precision of 86%, neutral of 1.00% and negative of 87%, positive recall of 90%, neutral 87% and negative of 26 % and positive F1-Score of 88% neutral 42% and negative 86%. Then at the Social Network Analysis stage there is the most influential account, namely influencer @alvinline21 with 1402 nodes.