Scholarlink

<p>In recent years, social media platforms like Facebook and Twitter have drastically reduced the degree of separation among users, accelerating the spread of information. While this enables rapid communication, it also facilitates the widespread propagation of fake news, posing serious challenges to society. Given the critical impact of misinformation on administrative and community decisions, accurate and timely fake news detection has become essential. This research paper addresses this problem by proposing HCA-FND (Hybrid Classification Approach for Fake News Detection), a novel hybrid two-step methodology. The first step employs classic machine learning algorithms to extract linguistic and stylistic features, while the second step utilizes deep transformer-based language models for context and sentiment analysis. Additionally, a real-time fact-checking module cross-references claims against reliable sources to improve accuracy. Extensive experiments demonstrate that HCA-FND outperforms existing methods by achieving a 3–5% improvement across key metrics–accuracy (96.3%), precision (95.8%), recall (94.9%), and F1-score (95.3%). Overall, HCA-FND offers a scalable, reliable, and interpretable solution for fake news detection.</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

Hca-fnd: a hybrid two-tiered approach for fake news detection using machine learning and natural language processing

  • 期刊论文

  • 发表日期：

    2026年1月6日

查看全文
  • V. Venkataramanan,
  • Anand Nayyar,
  • Pankaj Mishra,
  • Atharva Raut,
  • Vats S. Shah,
  • Vaibhav Vanage

摘要

In recent years, social media platforms like Facebook and Twitter have drastically reduced the degree of separation among users, accelerating the spread of information. While this enables rapid communication, it also facilitates the widespread propagation of fake news, posing serious challenges to society. Given the critical impact of misinformation on administrative and community decisions, accurate and timely fake news detection has become essential. This research paper addresses this problem by proposing HCA-FND (Hybrid Classification Approach for Fake News Detection), a novel hybrid two-step methodology. The first step employs classic machine learning algorithms to extract linguistic and stylistic features, while the second step utilizes deep transformer-based language models for context and sentiment analysis. Additionally, a real-time fact-checking module cross-references claims against reliable sources to improve accuracy. Extensive experiments demonstrate that HCA-FND outperforms existing methods by achieving a 3–5% improvement across key metrics–accuracy (96.3%), precision (95.8%), recall (94.9%), and F1-score (95.3%). Overall, HCA-FND offers a scalable, reliable, and interpretable solution for fake news detection.