Tired of clickbait-y and innacurate headlines, and long, boring articles in your news feed? The No-Clickbait Times is here to help. Using NLP, The No-Clickbait Times aims to analyse the headline and text, and summarise the article accordingly in a single line or two, giving an accurate description so your daily news feed can be precise and to the point.
numpy, pandas, matplotlib, seaborn
Scikit-Learn
gensim
nltk
GoogleNews-vectors-negative300.bin (training dataset, available online)
bert-extractive-summarizer
Run Final_model.ipynb with the requirements present to get an output csv that you can sift through using ResultComparision.ipynb.
Index.html is our sample webpage for displaying how the results could look with a finished product.
ClickbaitClassifier.ipynb, bertSummarizer.ipynb and SummaryReviewer.ipynb are the individual components that make up Final_model.ipynb
train.csv (labeled) and test.csv (unlabeled) are datasets extracted from kaggle.com/micdsouz/news-clickbait that we clean, train and demonstrate our model on.