Scholarlink

<p>Sarcasm detection has become an essential task in natural language processing, especially in social media, where sarcasm is frequently used to convey nonliteral meanings that can distort sentiment analysis outcomes. This study proposes a multimodal approach that integrates text and audio features to enhance sarcasm detection accuracy. Traditional sarcasm detection methods often rely solely on text-based features, which may fail to capture the vocal cues associated with sarcastic intent. Our proposed model combines text features such as TF-IDF, word embeddings, and sentiment polarity shifts with audio features, including pitch, energy intensity, and speech rate. By using a Support Vector Machine (SVM) classifier and optimizing the hyperparameters, we achieved an accuracy of 87.8%, significantly outperforming single-modality models. The results highlight the importance of multimodal fusion in sarcasm detection, as combining linguistic and prosodic features provides a more comprehensive understanding of sarcasm, especially in social media content, where users employ both text and vocal nuances to express sarcasm. This approach offers promising applications for enhancing sentiment analysis, social media monitoring, and conversational AI systems. Future work could explore additional modalities, such as visual elements in memes, and leverage deep learning models for further improvements.</p>

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

Sarcasm detection in social media text and audio conversation using SVM Classifier

  • 期刊论文

  • 发表日期：

    2026年7月7日

查看全文
  • Swati Tiwari,
  • Vivek Shukla,
  • Abhishek Shukla,
  • Rohit Miri,
  • Prakash Chandra Sharma,
  • Upasana Sinha,
  • Rohit Raja

摘要

Sarcasm detection has become an essential task in natural language processing, especially in social media, where sarcasm is frequently used to convey nonliteral meanings that can distort sentiment analysis outcomes. This study proposes a multimodal approach that integrates text and audio features to enhance sarcasm detection accuracy. Traditional sarcasm detection methods often rely solely on text-based features, which may fail to capture the vocal cues associated with sarcastic intent. Our proposed model combines text features such as TF-IDF, word embeddings, and sentiment polarity shifts with audio features, including pitch, energy intensity, and speech rate. By using a Support Vector Machine (SVM) classifier and optimizing the hyperparameters, we achieved an accuracy of 87.8%, significantly outperforming single-modality models. The results highlight the importance of multimodal fusion in sarcasm detection, as combining linguistic and prosodic features provides a more comprehensive understanding of sarcasm, especially in social media content, where users employ both text and vocal nuances to express sarcasm. This approach offers promising applications for enhancing sentiment analysis, social media monitoring, and conversational AI systems. Future work could explore additional modalities, such as visual elements in memes, and leverage deep learning models for further improvements.