2nd Position at Hackanova 3.0 TCET
FactFinder: An AI-powered fake news detection platform using BERT and ensemble learning models.
FactFinder – AI-Powered Fake News Detection Platform
Team: Yash G. Varma, Krishnan P.V., Sahil Brid, Sheshasai Dusa
Achievement: 2nd Position at Hackanova 3.0 TCET (36 hours)
Problem Statement
The proliferation of misinformation poses significant challenges to information integrity in the digital age. FactFinder addresses this by developing a web-based application that leverages machine learning to detect fake news.
Nature of the Solution
FactFinder is a web application designed to assess the authenticity of news articles. Users can input a news article, and the application processes the text using a pre-trained BERT model, which captures contextual nuances in the language. The ensemble model then classifies the article as either "Real" or "Fake."
Technologies Used
- ML Model: Ensemble model combining Bidirectional Encoder Representations from Transformers (BERT) and other classifiers.
- Language: Python
- Backend: Flask
- Frontend: HTML and CSS
- Dataset: WELFake dataset (72,000+ labeled news articles)