Analisis Sentimen Terhadap Cyberbullying Pada Komentar Di Media Sosial Tiktok Menggunakan Metode Tf-Idf Dan Logistic Regression

Ulumuddin, Ihya (2026) Analisis Sentimen Terhadap Cyberbullying Pada Komentar Di Media Sosial Tiktok Menggunakan Metode Tf-Idf Dan Logistic Regression. Undergraduate thesis, Politeknik Negeri Jember.

	Text (Abstrak) Abstrak.pdf - Submitted Version Available under License Creative Commons Attribution Share Alike. Download (10kB)
	Text (Bab 1 Pendahuluan) Bab 1 Pendahuluan.pdf - Submitted Version Available under License Creative Commons Attribution Share Alike. Download (55kB)
	Text (Daftar Pustaka) Daftar Pustaka.pdf - Submitted Version Available under License Creative Commons Attribution Share Alike. Download (10kB)
	Text (Laporan Lengkap) Laporan Lengkap.pdf - Submitted Version Restricted to Registered users only Download (1MB) | Request a copy

Abstract

This study aims to analyze the sentiment of TikTok comments related to cyberbullying using Term Frequency-Inverse Document Frequency (TF-IDF) and Logistic Regression. The classification is divided into two categories, positive and negative, where negative sentiment is not directly interpreted as cyberbullying. The dataset consists of 1,376 Indonesian-language comments collected through scraping, manually labeled, and validated by a psychology expert. The data processing stages include preprocessing, word weighting using TF-IDF, and sentiment classification using Logistic Regression. The dataset was divided into 1,100 training data and 276 testing data. Internal evaluation achieved an accuracy of 80.80%, with 223 correct predictions and 53 incorrect predictions. Additional testing on 150 comments achieved an accuracy of 70.67% and a macro-average F1-score of 51.71%, indicating that the model's performance in classifying both sentiment categories was unbalanced. The model was implemented as a website prototype using Flask and HTML, providing comment analysis through TikTok video links or direct text input. The results indicate that TF-IDF and Logistic Regression can be used for sentiment classification; however, the classification results are insufficient to directly determine the presence of cyberbullying.

Actions (login required)

View Item