Scholarlink

With the proliferation of social media, cyberbullying has emerged as a significant societal issue, necessitating robust detection mechanisms. This paper presents a novel data augmentation approach using Sentiment Analysis and WordNet-Based Comment Transformation (SAWCT), designed to enhance the performance of cyberbullying detection models. SAWCT leverages Aspect Sentiment Triplet Extraction (ASTE) task to identify aspect and opinion words within comments and employs WordNet to generate semantically rich augmentations. Results demonstrate the effectiveness of SAWCT across three Pre-trained Language Models (PLMs) and three datasets: HateXplain, OLID, and Cyberbullying Tweets (CT). The results indicate consistent improvements in precision, recall, and F1 score, with the most significant enhancements observed when combining aspect and opinion words. Ablation study further validates the importance of both components in improving detection accuracy. This work contributes a new perspective on leveraging sentiment analysis and WordNet to augment training data, enhance model understanding of varied and extensive expressions in cyberbullying contents.

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

A Data Augmentation Approach Using Sentiment Analysis and WordNet-Based Comments Transformation

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Ziyu Ye,
  • Yihan Wang,
  • Zhuowen Zhang,
  • Huakang Li

摘要

With the proliferation of social media, cyberbullying has emerged as a significant societal issue, necessitating robust detection mechanisms. This paper presents a novel data augmentation approach using Sentiment Analysis and WordNet-Based Comment Transformation (SAWCT), designed to enhance the performance of cyberbullying detection models. SAWCT leverages Aspect Sentiment Triplet Extraction (ASTE) task to identify aspect and opinion words within comments and employs WordNet to generate semantically rich augmentations. Results demonstrate the effectiveness of SAWCT across three Pre-trained Language Models (PLMs) and three datasets: HateXplain, OLID, and Cyberbullying Tweets (CT). The results indicate consistent improvements in precision, recall, and F1 score, with the most significant enhancements observed when combining aspect and opinion words. Ablation study further validates the importance of both components in improving detection accuracy. This work contributes a new perspective on leveraging sentiment analysis and WordNet to augment training data, enhance model understanding of varied and extensive expressions in cyberbullying contents.