This repository contains a deep learning-based pipeline designed to identify political leanings in news articles and social media text, and subsequently "flip" or neutralize that bias. By leveraging state-of-the-art Transformer architectures, the project aims to promote media literacy and provide a tool for consuming more balanced information.

The Newly Processed Dataset is again pre-processed where-- Tags, URLs, Noise, Duplicates are removed and the Dataset's Spelling are Checked(wrt Important Punctuations and Terminals) such that the Transformer can properly vectorize the Text.