Scholarlink

The rise of social media as a public space has led to increased cyberbullying, which can harm not only individuals but also brand reputation. This study aims to detect cyberbullying in Instagram comments using the Support Vector Machine (SVM) algorithm and to identify dominant discussion topics through Latent Dirichlet Allocation (LDA). The dataset consists of comments related to Ms. Glow and Dr. Richard Lee from August to December 2024, collected via manual web scraping. A quantitative-descriptive approach is applied, with comments classified into cyberbullying and non-cyberbullying categories. SVM achieves high accuracy in sentiment classification, while LDA reveals key themes such as deception, public distrust, and personal criticism. The findings contribute to early cyberbullying detection and offer insights for digital communication strategy and brand reputation management.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Cyberbullying Detection Framework Using Support Vector Machine and Latent Dirichlet Allocation: Analyzing Instagram Comments for Brand Reputation Insights

  • 章节

  • 发表日期：

    2026年

查看全文
  • Suwanti Dean,
  • Yusza Reditya Murti,
  • Nurvita Trianasari

摘要

The rise of social media as a public space has led to increased cyberbullying, which can harm not only individuals but also brand reputation. This study aims to detect cyberbullying in Instagram comments using the Support Vector Machine (SVM) algorithm and to identify dominant discussion topics through Latent Dirichlet Allocation (LDA). The dataset consists of comments related to Ms. Glow and Dr. Richard Lee from August to December 2024, collected via manual web scraping. A quantitative-descriptive approach is applied, with comments classified into cyberbullying and non-cyberbullying categories. SVM achieves high accuracy in sentiment classification, while LDA reveals key themes such as deception, public distrust, and personal criticism. The findings contribute to early cyberbullying detection and offer insights for digital communication strategy and brand reputation management.