Fake News Detection on Online Social Network Using Machine Learning

Abstract

The rapid growth of online social networks has increased the spread of fake news, creating challenges in maintaining trustworthy information. Fake news can influence public opinion, create misinformation, and negatively impact society. Traditional methods of identifying fake news rely on manual verification, which is time-consuming and inefficient for large volumes of content. This project proposes a Fake News Detection System using Machine Learning techniques to automatically classify news as real or fake. The system processes textual content through preprocessing techniques such as tokenization, stop-word removal, and feature extraction. Machine learning algorithms are then applied to detect patterns associated with misleading information. The model is trained and tested on news datasets and demonstrates effective performance in identifying fake news. The proposed system provides faster and more reliable detection, helping users verify online information and reduce the spread of misinformation.

Introduction

Online social networks have become major platforms for sharing information, but they have also enabled the rapid spread of fake news. Fake news consists of false or misleading information presented as genuine news, which can influence public opinion, political decisions, social behavior, and trust. Traditional fact-checking methods rely on human experts and media organizations, but they are slow, labor-intensive, and unable to handle the vast amount of online content.

To address this issue, the proposed Fake News Detection System uses Machine Learning (ML) and Natural Language Processing (NLP) to automatically identify fake news. NLP techniques analyze textual content, linguistic patterns, and contextual information, while ML algorithms classify news as real or fake with high accuracy.

Methodology

The system follows several stages:

1. Data Collection – Gathering labeled datasets containing real and fake news articles.
2. Text Preprocessing – Cleaning the data through:
   • Lowercase conversion
   • Tokenization
   • Stop-word removal
   • Stemming
   • Lemmatization
   • Removal of special characters
3. Feature Extraction – Converting text into numerical representations using:
   • Bag of Words (BoW)
   • Term Frequency–Inverse Document Frequency (TF-IDF)
4. Model Training – Training machine learning models such as:
   • Logistic Regression
   • Naïve Bayes
   • Decision Trees
   • Random Forest
   • Support Vector Machine (SVM)
5. Prediction and Evaluation – Testing the model on unseen data and classifying news articles as real or fake.

System Implementation

• Frontend: Developed using HTML, CSS, JavaScript, and React.js for a user-friendly interface.
• Backend: Built with Python and Flask for data processing and communication.
• Machine Learning Tools: Implemented using Scikit-learn, TensorFlow, and NLP libraries.

Users can enter news content, and the system processes the text and displays the prediction result along with a confidence score.

Results

The proposed system achieved strong performance:

• Accuracy: ~95%
• Precision: 94%
• Recall: 93%
• F1-Score: High overall classification effectiveness

The model successfully identified both fake and real news articles, demonstrating that ML and NLP techniques can effectively combat misinformation on online platforms.

Conclusion

The proposed Fake News Detection System successfully demonstrates an automated and intelligent approach for identifying misleading information on online social networks using Machine Learning and Natural Language Processing techniques. The system processes textual content, extracts meaningful features, and classifies news articles into real or fake categories with high efficiency and reliability. The developed model was trained and evaluated using labelled datasets and achieved satisfactory classification performance with high accuracy. By reducing dependency on manual verification methods, the proposed approach improves detection speed and enables large-scale content analysis. The use of machine learning algorithms allows the system to recognize hidden textual patterns and make accurate predictions. In conclusion, the developed system provides an effective solution for minimizing the spread of fake news and improving trust in online information sharing. The project establishes a strong foundation for future improvements in real-time misinformation detection and intelligent social media monitoring systems.

References

[1] K. Shu, A. Sliva, S. Wang, J. Tang, and H. Liu, “Fake News Detection on Social Media: A Data Mining Perspective,” ACM SIGKDD Explorations Newsletter, vol. 19, no. 1, pp. 22–36, 2017. [2] H. Ahmed, I. Traore, and S. Saad, “Detection of Online Fake News Using Machine Learning Techniques,” International Conference on Intelligent Systems, 2018. [3] J. Wang, “LIAR: A Benchmark Dataset for Fake News Detection,” Proceedings of ACL, 2017. [4] Y. Goldberg, “A Primer on Neural Network Models for Natural Language Processing,” Journal of Artificial Intelligence Research, 2016. [5] TensorFlow Documentation, “https://www.tensorflow.org”, Accessed 2026. [6] Kaggle Fake News Dataset, “https://www.kaggle.com”, A

Copyright

Copyright © 2026 Ambrish ., Abhishek D G, Aditya , Nithesh , Prof. Suresh Kumar Biradar . This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.