Fake News Detector

• 

  AI Model (Input)

• 

  Code of Final Result

• 

  Output Cell

• 

  Dataset for Final Result

• 

  Final Visualization

• 

  Visualization Process

🎯 Inspiration:

This project aims to evaluate the effectiveness of Artificial Intelligence in detecting fake news and compare its accuracy with human intelligence. Using a pre-trained model from Hugging Face, we analyzed both real and fake news articles, while human participants submitted their predictions through Google Forms. The project highlights how AI and human reasoning align or differ when it comes to judging the authenticity of news.

❓ What It Does:

Collected 12 news articles (6 real + 6 fake)

Ran each article through an AI model for predictions

Collected 20+ human responses using Google Forms

Compared AI predictions vs human responses

Visualized accuracy and insights using Excel charts

Shared findings via a demo video and GitHub repository

🛠️ How We Built It?

Python (developed using Google Colab)

Hugging Face Transformers for text classification

Pre-trained Model: vikram71198/distilroberta-base-finetuned-fake-news-detection

Google Forms for human predictions

Microsoft Excel for survey data and visualizations

GitHub for project hosting

YouTube for demo video

🚧 Challenges We Faced:

Finding unbiased news articles that weren’t clearly labeled

Selecting a balanced and reliable AI model

Designing a user-friendly Google Form

Formatting AI predictions with confidence scores

Ensuring diverse participation within the deadline

🏆 Accomplishments We’re Proud Of:

Successfully demonstrated a side-by-side comparison of AI vs Human

Built a complete workflow from news collection to result visualization

Created a demo video and hosted project on GitHub

👩‍💻 About the Team — The A_Coders:

We are two passionate learners: Asima and Areeba — and we proudly call ourselves The A_Coders.

📌 Task Division Asima: News collection, cleanup, AI model implementation, Excel visualizations, GitHub uploads, and overall submission.

Areeba: Form design, survey distribution, response handling, demo video recording, and manual data entry.

Both worked together on content refinement, interpreting predictions, and preparing the final presentation.

📚 What We Learned:

This project goes beyond fake news detection — it explores the reliability of AI vs human judgment. We gained insight into their strengths, weaknesses, and how both can complement each other in today’s digital age.

🔮 What’s Next for Fake News Detector:

Use more advanced, multilingual AI models

Expand dataset to 50+ articles

Automate AI-human comparison via dashboard

Collaborate with media literacy organizations

🔗 Useful Links:

Google Forms: (https://docs.google.com/forms/d/e/1FAIpQLSepa571pwyVrF4cks-d4ynOSe-bdTCMwMnGknElC5r-SNYEPg/viewform?usp=header) , (https://docs.google.com/forms/d/e/1FAIpQLSfW03-C-XeR7VtFydFlVzzkF1_rtbiwUJKW850nlzYXwQUm5g/viewform?usp=header)

YouTube Demo Video: (https://youtu.be/jYQHRCI76ZE?si=2Ea_eh0jLHsV62m3)

GitHub Repository: (https://github.com/AsimaShafiq/Fake_News_Detector/tree/main)

Built With

• excel
• fake-news-detection-model
• google-colab
• google-docs
• hugging-face-transformer-library
• python
• video-editor