Web Search Engine Misinformation Notifier Extension (SEMiNExt): A Machine Learning Based Approach during COVID-19...

Area of Science:

• Health Informatics
• Computational Linguistics
• Public Health

Background:

• Misinformation regarding diseases like COVID-19, treatments, and vaccines poses significant public health risks.
• Existing surveillance tools struggle to combat the rapid online spread of unverified health information in real-time.
• There is a critical need for immediate tools to verify health information during online searches.

Purpose of the Study:

• To develop and evaluate a real-time web search engine misinformation notifier extension (SEMiNExt).
• To bridge the gap in immediate, user-facing misinformation detection during online health inquiries.
• To enhance the safety and reliability of online health information seeking.

Main Methods:

• Integration of Natural Language Processing (NLP) and machine learning algorithms into a browser extension.
• Real-time classification of the veracity of user search queries.
• Development of a notification system to alert users about the authenticity of health information.

Main Results:

• SEMiNExt achieved high performance metrics using an Artificial Neural Network (ANN) model.
• The system demonstrated 93% accuracy, 92% F1-score, 92% precision, and 93% recall with 80% data training.
• ANN models showed high predictive accuracy even with limited training data, crucial for early detection.

Conclusions:

• SEMiNExt effectively identifies and notifies users about potential health misinformation in real-time.
• The tool significantly improves online health management by enabling safer web searching.
• Early detection of emerging misinformation via SEMiNExt can mitigate public health risks and enhance safety.