Scholarlink

Rumors spread quickly on social media, which profoundly affects people’s ideas and how they think about themselves and society. Identifying the rumors and refuting them before they cause suffering and agony is essential. This work presents the techniques for detecting rumors, especially on social media, to stop stress and injury of any kind. Feature extraction is done using the hybrid approach to choose the relevant features. Sentiment analysis is used in conjunction with machine learning to detect rumors. The standard PHEME dataset is used. Several classification techniques including Multinomial Naïve Bayes, Logistic Regression, Random Forest, Decision Tree, and K-Nearest Neighbor are implemented to compare the accuracy of obtained results. The Multinomial Bayes algorithm provides better results as compared to other classification approaches. If it is a rumor then it specifies whether it is harmful or not, as a harmful rumor affects society.

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

An Additive Approach Using Machine Learning and Sentimental Analysis for Pernicious Rumor Detection in the Social Media

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Priyanshi Borase,
  • Satish Kolhe

摘要

Rumors spread quickly on social media, which profoundly affects people’s ideas and how they think about themselves and society. Identifying the rumors and refuting them before they cause suffering and agony is essential. This work presents the techniques for detecting rumors, especially on social media, to stop stress and injury of any kind. Feature extraction is done using the hybrid approach to choose the relevant features. Sentiment analysis is used in conjunction with machine learning to detect rumors. The standard PHEME dataset is used. Several classification techniques including Multinomial Naïve Bayes, Logistic Regression, Random Forest, Decision Tree, and K-Nearest Neighbor are implemented to compare the accuracy of obtained results. The Multinomial Bayes algorithm provides better results as compared to other classification approaches. If it is a rumor then it specifies whether it is harmful or not, as a harmful rumor affects society.