Hi! I’m new to all this AI stuff and currently taking the fast.ai course which is great.
As a study project, I decided to create a model that makes pros and cons list from the customer reviews.
My question is is it a good idea to finetune the text summarization model (e.g. “bart-large-cnn”) to get a list of characteristics of a product and then use a classifier to decide if each point of the list is positive or negative? Or maybe there are much better ways to do so?
Maybe someone already did something similar and could point out some reference articles, papers, or even notebooks?