Abstract: With the increasing popularity of social media platforms like YouTube, monitoring and regulating user-generated content has become a significant challenge. In particular, detecting hate speech and hate crimes within YouTube comments is a critical task to ensure user safety, foster inclusive online communities, and comply with legal regulations.

This abstract presents an overview of a research study focused on hate crime detection in YouTube comments using machine learning (ML) techniques. The objective of the study is to develop an automated system capable of identifying and flagging hate speech and potential hate crimes within user comments.

The research involves collecting a large dataset of YouTube comments labeled for hate speech, hate crimes, and non-offensive content. Our work encompasses two main contributions. Firstly, we have developed a detailed taxonomy for classifying hateful online comments, considering both the types of hate speech and the targets of such comments. This taxonomy enables a more granular understanding and analysis of hate speech occurrences in online social media.

Secondly, we have conducted an extensive machine learning experiment using various algorithms, including Logistic Regression, Decision Tree, Random Forest, Adaboost, and Linear SVM. The goal was to create a multiclass, multilabel classification model capable of automatically detecting and categorizing hateful comments within the realm of online social media.

To ensure the reliability of our model, we performed validation tests to assess its predictive capability. Additionally, this research has provided valuable insights into the distinct types of hate speech prevalent on social media platforms.

Keywords: Hate, YouTube, social media, offensive, Muslim, jihad, fool.

PDF | DOI: 10.17148/IJARCCE.2023.12671

Open chat
Chat with IJARCCE