How to make a reference book from the HTML pages of a Q&A forum?

Question

Parsing

What is given: subject, content (rough outline of the book, structure by subtopics).
Required: to fill the content from HTML files (files: text and images - already available locally) so that you get a structured guide to the most popular and interesting issues, arranged according to the specified content structure.
At the same time, it is necessary, as accurately as possible in a semi-automatic mode, to exclude any answers and comments that are not related to the question.
How can this be done?
Thank you!

Answer 1

1. Ask Google a specific question and make up the first N-links, the model of the correct answer - train the model on the search results.
2. Apply the trained model to the current question and find the most appropriate answer (from all the answers and comments).

Answer 2

Forum2Book converter, but if you are serious write your own parser or look for freelancers

Answer 3

there is no way to do it until they come up with an AI that will weed out all sorts of answers and comments for you