Analisis Sentimen Trend Fashion Pada Sosial Media Instagram Di Indonesia Untuk Mengetahui Opini Publik Dengan Metode Naive Bayes

Khalishah, Syarafina (2026) Analisis Sentimen Trend Fashion Pada Sosial Media Instagram Di Indonesia Untuk Mengetahui Opini Publik Dengan Metode Naive Bayes. Undergraduate thesis, Politeknik Negeri Jember.

	Text (Ringkasan/Abstract) Ringkasan⁄Abstract.pdf - Submitted Version Available under License Creative Commons Attribution Share Alike. Download (136kB)
	Text (Bab 1 Pendahuluan) Bab 1 Pendahuluan.pdf - Submitted Version Available under License Creative Commons Attribution Share Alike. Download (201kB)
	Text (Daftar Pustaka) Daftar Pustaka.pdf - Submitted Version Available under License Creative Commons Attribution Share Alike. Download (111kB)
	Text (Laporan Lengkap) Laporan Lengkap.pdf - Submitted Version Restricted to Registered users only Download (4MB) | Request a copy

Abstract

The rapid development of social media, particularly Instagram, has transformed fashion consumption patterns in Indonesia, making online consumer opinions a key determinant of brand perception and purchasing decisions. This study applies sentiment analysis to classify public opinion toward two prominent local Indonesian fashion brands, This Is April and MyRubylicious, using 1,813 Instagram comments (1,013 from This Is April and 800 from MyRubylicious) obtained through web scraping. The data were processed through a series of preprocessing stages including cleaning, tokenizing, normalizing, stopword removal, and stemming, followed by feature weighting using TF-IDF before being classified with the Naïve Bayes algorithm using a data split ratio of 70% training and 30% testing. Model evaluation was conducted using a confusion matrix measuring accuracy, precision, recall, and F1-Score. The evaluation results show that the model for This Is April achieved an accuracy of 98.36%, precision of 98.73%, recall of 98.11%, and F1- Score of 98.42%, while MyRubylicious achieved an accuracy of 99.17% with precision, recall, and F1-Score each at 99.21%. Sentiment classification on both brands was dominated by positive comments, reflecting strong consumer acceptance of both brands on Instagram. These findings demonstrate the effectiveness of the Naïve Bayes algorithm in classifying Indonesian-language sentiment, while also generating data-driven insights that local fashion industry players can leverage to design more targeted and consumer-responsive marketing strategies.

Actions (login required)

View Item