A large chemical company keeps its employees up to date on news by compiling a list of news articles to send to its staff. Recently, its news feed has been infiltrated by fake news articles. I used AI and Natural Language Processing to build a classifier to separate the genuine articles from the fake news articles.
I build a Naive Bayes model that was 99% accurate in predicting real and fake news. I used over 600 real and fake news articles to train and validate the model.
- Modules - A custom cleaning module for preparing news articles for modeling.
- Notebooks - Contains three Jupyter Notebooks that I used to scrape news articles, read and clean the articles, and build a model.
- Saved Files - Intermediate files to save my progress