Identify pros and cons of aspect in the review text

Hi.
I have a dataset that looks like this:

What I’m trying to achieve is finding a way to extract pros and cons lists from product reviews. I tried to fine-tune BART Large Model based on this blog post. But the result was bad. While some proses and conses were related to the specified product, the part of them wasn’t. Also the other products (that occurred in the review text for comparison) popups in the results.

Does anybody know some good resources on the task of text summarization, or maybe there are some other approaches that suit this task more?

Any ideas on how to proceed? I will gladly accept any help! Thank you!