BART-based article summarization with TF-IDF sentence ranking, evaluated using ROUGE, METEOR, and BERTScore metrics.
Develop upon the learning of our previous showcase, and actually use a natural language model (pre-trained) for article summarization.
Try to keep a simple code structure so that we can define a function out of it and use it for more than one article.
The beginning steps will closely mirror the ones from keyword extractor, but will diverge later on, as we’ll focus more on textual summarization rather than visualization.
The full notebook file will contain comments regarding our process, but we want to provide additional context and a visual breakdown. Where necessary, output will be provided as captions!