Fake social media news and distorted campaign detection framework using sentiment analysis & machine learning

Area of Science:

• Computer Science
• Artificial Intelligence
• Natural Language Processing

Background:

• Social media platforms facilitate global information exchange but are increasingly exploited for spreading fake news and spam.
• Manual verification of vast amounts of social media content is infeasible, necessitating automated solutions.
• Existing methods struggle to accurately distinguish genuine content from disinformation and malicious bot activity.

Discussion:

• This research proposes a framework leveraging sentiment analysis based on emotions to analyze social media content.
• The model computes sentiment scores for content entities to identify fake or spam profiles and bot accounts.
• Sentiment analysis provides a nuanced approach to understanding user emotions and detecting deceptive content.

Key Insights:

• The developed framework effectively detects fake news, spam, and bot accounts with high precision.
• Sentiment analysis based on emotions proves to be a powerful tool for uncovering disinformation campaigns.
• The machine learning algorithm achieved a remarkable accuracy of 99.68%, significantly surpassing other methodologies.

Outlook:

• Further research can explore incorporating multimodal sentiment analysis (text, image, video) for enhanced detection.
• The framework has the potential to be integrated into social media platforms to improve content moderation.
• Continued development of AI-driven solutions is crucial for combating the growing threat of online disinformation.