Comment Sentiment Analysis of JNE Using K-Nearest Neighbor (KNN) Method on Twitter

Abstract

Social media has evolved into a prominent public space for virtual criticism, particularly on platforms like Twitter, facilitated by widespread smartphone usage. Netizens utilize Twitter as an effective communication channel due to its accessibility and vast reach. This study focuses on sentiment analysis of comments from the public on Twitter, aiming to expedite the acquisition of accurate information about the general sentiment towards JNE (a logistics company). The K-Nearest Neighbor (KNN) classifier is employed, employing the TF-IDF weighting method to classify Indonesian language comments and assess the achieved accuracy. Highlights: Study focused on sentiment analysis of Twitter comments concerning JNE services using the K-Nearest Neighbor (KNN) method with Indonesian language text. Employed the TF-IDF weighting to classify comments and achieved an impressive 90% accuracy in sentiment analysis. The obtained classification proves valuable in evaluating public perception of JNE's services based on feedback from the social media community on Twitter.