Leveraging Natural Language Processing for Automated Misinformation Detection in Online News

  • Muhammad Hammad u Salam
  • Shujaat Ali Rathore
  • Muhammad Irfan
  • Kanwal Ameen
  • Muhammad Atif

Abstract

The rise of social media and online news has resulted in a dramatic increase in fake news and misinformation which poses serious challenges to public trust and information credibility. These old approaches to fact checking are unable to match the speed at which misleading information is shared making more more automated techniques necessary. This research seeks to develop techniques to identify online articles containing misinformation using NLP (natural language processing). The approach uses transformational deep learning models like BERT and RoBERTA, machine learning based text classifiers, and sentiment analysis tools to evaluate the truthfulness of online news articles. Furthermore, semantic similarity measures, stance detection, and linguistic feature measurements are used to separate lies from the truth. Experimental results suggest that NLP-based fake news detectors perform better than traditional systems based of rules and keywords and provide quality results in terms of accuracy, precision, and recall. The study advocates for the use of AI based systems as an automated tool to help fight against the spread of false information online.

Published
2024-09-15